Diagnostic Ultrasound, 2-Volume Set, 4th Edition

Citation preview

DIAGNOSTIC ULTRASOUND FOURTH EDITION

Carol M. Rumack, MD, FACR Professor of Radiology and Pediatrics University of Colorado Denver School of Medicine Denver, Colorado

Stephanie R. Wilson, MD, FRCPC Clinical Professor of Radiology University of Calgary Staff Radiologist Foothills Medical Centre Calgary, Alberta, Canada

J. William Charboneau, MD, FACR Professor of Radiology Mayo Clinic College of Medicine Consultant in Radiology Mayo Clinic Rochester, Minnesota

Deborah Levine, MD, FACR Professor of Radiology Harvard Medical School Associate Radiologist-in-Chief of Academic Affairs Director of Ob/Gyn Ultrasound Beth Israel Deaconess Medical Center Boston, Massachusetts

1600 John F. Kennedy Blvd. Ste 1800 Philadelphia, PA 19103-2899

DIAGNOSTIC ULTRASOUND, FOURTH EDITION Copyright © 2011 by Mosby, Inc., an affiliate of Elsevier Inc.

ISBN: 978-0-323-05397-6

No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher. Details on how to seek permission, further information about the Publisher’s permissions policies and our arrangements with organizations such as the Copyright Clearance Center and the Copyright Licensing Agency, can be found at our website: www.elsevier.com/permissions. This book and the individual contributions contained in it are protected under copyright by the Publisher (other than as may be noted herein).

Notices Knowledge and best practice in this field are constantly changing. As new research and experience broaden our understanding, changes in research methods, professional practices, or medical treatment may become necessary. Practitioners and researchers must always rely on their own experience and knowledge in evaluating and using any information, methods, compounds, or experiments described herein. In using such information or methods they should be mindful of their own safety and the safety of others, including parties for whom they have a professional responsibility. With respect to any drug or pharmaceutical products identified, readers are advised to check the most current information provided (i) on procedures featured or (ii) by the manufacturer of each product to be administered, to verify the recommended dose or formula, the method and duration of administration, and contraindications. It is the responsibility of practitioners, relying on their own experience and knowledge of their patients, to make diagnoses, to determine dosages and the best treatment for each individual patient, and to take all appropriate safety precautions. To the fullest extent of the law, neither the Publisher nor the authors, contributors, or editors, assume any liability for any injury and/or damage to persons or property as a matter of products liability, negligence or otherwise, or from any use or operation of any methods, products, instructions, or ideas contained in the material herein.

Previous editions copyrighted 2005, 1998, 1993 by Mosby, Inc.

Library of Congress Cataloging in Publication Data Diagnostic ultrasound / [edited by] Carol M. Rumack … [et al.].—4th ed.    p. ; cm   Includes bibliographical references and index.   ISBN 978-0-323-05397-6 (hardcover : alk. paper)  1.  Diagnostic ultrasonic imaging.  I.  Rumack, Carol M.   [DNLM:  1.  Ultrasonography. WN 208]   RC78.7.U4D514 2011   616.07′543—dc22 2010034851

Acquisitions Editor: Rebecca Gaertner Developmental Editor: Lisa Barnes Publishing Services Manager: Patricia Tannian Team Manager: Radhika Pallamparthy Senior Project Manager: John Casey Project Manager: Anitha Sivaraj Designer: Steven Stave

Working together to grow libraries in developing countries www.elsevier.com | www.bookaid.org | www.sabre.org

Printed in United States of America Last digit is the print number: 9  8  7  6  5  4  3  2  1 

About the Editors Carol M. Rumack, MD, is Professor of Radiology and Pediatrics at the University of Colorado Denver School of Medicine in Denver, Colorado. Her clinical practice is based at the University of Colorado Hospital. Her primary research has been in neonatal sonography of high-risk infants, particularly the brain. Dr. Rumack has published widely in this field and lectured frequently on pediatric ultrasound. She is a fellow and past president of the American College of Radiology, a fellow of both the American Institute of Ultrasound in Medicine and the Society of Radiologists in Ultrasound. She and her husband, Barry, have two children, Becky and Marc.

Stephanie R. Wilson, MD, is Clinical Professor of Radiology at the University of Calgary where she heads a specialty ultrasound clinic at the Foothills Medical Centre devoted primarily to the imaging of diseases of the gastrointestinal tract and gynecologic organs. With support from the Canadian Institute of Health Research (CIHR), Dr. Wilson worked with Dr. Peter Burns in Toronto on the characterization and detection of focal liver masses with contrastenhanced ultrasound (CEUS) and is an established authority in this field. A recognized expert on ultrasound of the gastrointestinal tract and abdominal and pelvic viscera, she is the recipient of many university teaching awards and is a frequent international speaker and author. Dr. Wilson was the first woman president of the Canadian Association of Radiologists (CAR) and is the current president-elect of the International Contrast Ultrasound Society (ICUS). She has received the gold medal from CAR in recognition of her contribution to radiology. A golf enthusiast, she and her husband Ken, have two children, Jessica and Jordan.

J. William Charboneau, MD, is Professor of Radiology at the Mayo Clinic in Rochester, Minnesota. His current research interests include image-guided tumor biopsy and ablation, as well as sonography of the liver and small parts. He is coauthor of over 200 publications, assistant editor of the Mayo Clinic Family Health Book, and an active lecturer nationally and internationally. He is a fellow in the American College of Radiology and the Society of Radiologists in Ultrasound. He and his wife, Cathy, have three children, Nick, Ben, and Laurie. Deborah Levine, MD, is Professor of Radiology at Beth Israel Deaconess Medical Center, Boston, and Harvard Medical School. At Beth Israel Deaconess Medical Center she is Associate Radiologist-in-Chief of Academic Affairs, Co-Chief of Ultrasound, and Director of Ob/Gyn Ultrasound. Her main areas of clinical interest are obstetric and gynecologic ultrasound. Her research has focused on fetal magnetic resonance imaging as an aid to improving ultrasound diagnosis. Dr. Levine is an American College of Radiology Chancellor, Chair of the American College of Radiology Commission on Ultrasound, a fellow of the American Institute of Ultrasound in Medicine and Society of Radiologists in Ultrasound. She and her husband, Alex, have two children, Becky and Julie.

Contributors Jodi F. Abbott, MD Associate Professor Boston University School of Medicine Director of Antenatal Testing Boston Medical Center Boston, Massachusetts

Daryl J. Barth, RVT, RDMS Ultrasound Assistant Department of Sonography OSI St. Francis Medical Center Ultrasound Assistant Central Illinois Radiological AssociatesPeoria, Illinois

Jacques S. Abramowicz, MD, FACOG Frances T. & Lester B. Knight Professor Rush University Director, Ob/Gyn Ultrasound Rush University Medical Center Co-Director, Rush Fetal and Neonatal Medicine Program Rush University Chicago, Illinois

Beryl Benacerraf, MD Clinical Professor of Obstetrics and Gynecology and Radiology Brigham and Women’s Hospital Massachusetts General Hospital Harvard Medical School Boston, Massachusetts

Ronald S. Adler, PhD, MD Professor of Radiology Weill Medical College of Cornell University Chief, Division of Ultrasound and Biology Imaging Department of Radiology and Imaging Hospital for Special Surgery Attending Radiologist Department of Radiology New York Presbyterian Hospital New York City, New York Amit R. Ahuja, MD Diagnostic Imaging Resident Foothills Medical Centre Calgary, Alberta, Canada Jean M. Alessi-Chinetti, BS, RDMS, RVT Technical Director Vascular Laboratory Tufts Medical Center Boston, Massachusetts Thomas Atwell, MD Assistant Professor of Radiology Mayo Clinic College of Medicine Consultant in Radiology Mayo Clinic Rochester, Minnesota Diane S. Babcock, MD Professor of Radiology and Pediatrics University of Cincinnati College of Medicine Professor of Radiology and Pediatrics Cincinnati Children’s Hospital Medical Center Cincinnati, Ohio Carol E. Barnewolt, MD Assistant Professor of Radiology Harvard Medical School Director, Division of Ultrasound Children’s Hospital Boston Boston, Massachusetts

Carol B. Benson, MD Professor of Radiology Harvard Medical School Director of Ultrasound and Co-Director of High Risk Obstetrical Ultrasound Brigham and Women’s Hospital Boston, Massachusetts Raymond E. Bertino, MD, FACR, FSRU Medical Director of Vascular and General Ultrasound OSF Saint Francis Medical Center Clinical Professor of Radiology and Surgery University of Illinois College of Medicine Peoria, Illinois Edward I. Bluth, MD, FACR Clinical Professor Tulane University School of Medicine Chairman Emeritus Radiology Ochsner Health System New Orleans, Louisiana J. Antonio Bouffard, MD Senior Staff Radiologist Henry Ford Hospital Detroit, Michigan Consultant Radiologist James Andrews Orthopedics and Sports Medicine Center Pensacola, Florida Bryann Bromley, MD Clinical Associate Professor of Obstetrics and Gynecology Massachusetts General Hospital Clinical Associate Professor of Obstetrics and Gynecology and Radiology Brigham and Women’s Hospital Boston, Massachusetts

vi    Contributors Dorothy I. Bulas, MD Professor of Radiology and Pediatrics George Washington University Medical Center Pediatric Radiologist Children’s National Medical Center Washington, District of Columbia Peter N. Burns, PhD Professor and Chairman Department of Medical Biophysics University of Toronto Senior Scientist Department of Imaging Research Sunnybrook Health Sciences Centre Toronto, Ontario, Canada Barbara A. Carroll, MD Professor Emeritus of Radiology Department of Radiology Duke University Medical Center Durham, North Carolina J. William Charboneau, MD, FACR Professor of Radiology Mayo Clinic College of Medicine Consultant in Radiology Mayo Clinic Rochester, Minnesota Humaira Chaudhry, MD Fellow in Abdominal Imaging Duke University Medical Center Durham, North Carolina Tanya P. Chawla, MD, FRCPC Assistant Professor University of Toronto Toronto, Ontario, Canada David Chitayat, MD, FABMG, FACMG, FCCMG, FRCPC Professor University of Toronto Prenatal Diagnosis and Medical Genetics Program Department of Obstetrics and Gynecology Mount Sinai Hospital Toronto, Ontario, Canada Peter L. Cooperberg, MD Chief of Radiology St. Paul’s Hospital Chief of Radiology University of British Columbia Vancouver, British Columbia, Canada Peter M. Doubilet, MD, PhD Professor of Radiology Harvard Medical School Senior Vice Chair Department of Radiology Brigham and Women’s Hospital Boston, Massachusetts

Julia A. Drose, BA, RDMS, RDCS, RVT Associate Professor of Radiology University of Colorado at Denver Health Sciences Center Chief Sonographer Divisions of Ultrasound and Prenatal Diagnosis & Genetics University of Colorado Hospital Aurora, Colorado Beth S. Edeiken-Monroe, MD Professor of Radiology Department of Diagnostic Radiology The University of Texas Houston Medical School MD Anderson Cancer Center Houston, Texas Judy Estroff, MD Associate Professor of Radiology Harvard Medical School Division Chief, Fetal Neonatal Radiology Children’s Hospital Boston Radiologist Department of Radiology Beth Israel Deaconess Medical Center Radiologist Department of Radiology Brigham and Women’s Hospital Boston, Massachusetts Amy Symons Ettore, MD Consultant Department of Radiology Mayo Clinic College of Medicine Rochester, Minnesota Katherine W. Fong, MBBS, FRCPC Associate Professor of Medical Imaging and Obstetrics and Gynecology University of Toronto Faculty of Medicine Co-director, Centre of Excellence in Obstetric Ultrasound Mount Sinai Hospital Toronto, Ontario; Canada Bruno D. Fornage, MD Professor of Radiology and Surgical Oncology M. D. Anderson Cancer Center Houston, Texas J. Brian Fowlkes, PhD Associate Professor University of Michigan Department of Radiology Ann Arbor, Michigan Phyllis Glanc, MDCM Assistant Professor Department of Medical Imaging University of Toronto Assistant Professor Department of Obstetrics & Gynecology University of Toronto Site Director Body Imaging Women’s College Hospital Toronto, Ontario, Canada

Contributors    vii

Brian Gorman, MB, BCh, FRCR, MBA Assistant Professor of Radiology Mayo Clinic College of Medicine Consultant in Radiology Mayo Clinic Rochester, Minnesota S. Bruce Greenberg, MD Professor University of Arkansas for Medical Sciences Professor Arkansas Children’s Hospital Little Rock, Arkansas Leslie E. Grissom, MD Clinical Professor of Radiology and Pediatrics Department of Radiology Thomas Jefferson Medical College Thomas Jefferson University Hospital Philadelphia, Pennsylvania; Chair, Medical Imaging Department Medical Imaging Department—Radiology Alfred I. DuPont Hospital for Children Wilmington, Delaware; Pediatric Radiologist Medical Imaging Department—Radiology Christiana Care Health System Newark, Delaware Benjamin Hamar, MD Instructor of Obstetrics, Gynecology, and Reproductive Biology Beth Israel–Deaconess Medical Center Boston, Massachusetts Anthony E. Hanbidge, MB, BCh, FRCPC Associate Professor University of Toronto Head, Division of Abdominal Imaging University Health Network Mount Sinai Hospital and Women’s College Hospital Toronto, Ontario, Canada H. Theodore Harcke, MD, FACR, FAIUM Professor of Radiology and Pediatrics Jefferson Medical College Philadelphia, Pennsylvania Chief of Imaging Research Department of Medical Imaging Alfred I. DuPont Hospital for Children Wilmington, Delaware Ian D. Hay, MD Professor of Medicine Dr. R. F. Emslander Professor in Endocrinology Research Division of Endocrinology and Internal Medicine Mayo Clinic Consultant in Endocrinology and Internal Medicine Department of Medicine Mayo ClinicRochester, Minnesota Christy K. Holland, PhD Professor Departments of Biomedical Engineering and Radiology University of Cincinnati Cincinnati, Ohio

Caroline Hollingsworth, MD Assistant Professor of Radiology Duke University Medical Center Durham, North Carolina Bonnie J. Huppert, MD Assistant Professor of Radiology Mayo Clinic College of Medicine Consultant in Radiology Mayo Clinic Rochester, Minnesota E. Meridith James, MD, FACR Professor of Radiology Mayo Clinic College of Medicine Consultant in Radiology Mayo Clinic Rochester, Minnesota Susan D. John, MD Professor of Radiology and Pediatrics Chair, Department of Diagnostic and Interventional Imaging University of Texas Medical School at Houston Houston, Texas Neil D. Johnson, MBBS, MMed, FRANZCR Professor, Radiology and Pediatrics Cincinnati Children’s Hospital Medical Center Cincinnati, Ohio Korosh Khalili, MD, FRCPC Assistant Professor University of Toronto Staff Radiologist University Health Network Toronto, Ontario, Canada Beth M. Kline-Fath, MD Assistant Professor of Radiology Cincinnati Children’s Hospital Medical Center Cincinnati, Ohio Clifford S. Levi, MD, FRCPC Section Head Health Sciences Centre Professor University of Manitoba Winnipeg, Manitoba, Canada Deborah Levine, MD, FACR Professor of Radiology Harvard Medical School Associate Radiologist-in-Chief of Academic Affairs Director of Ob/Gyn Ultrasound Beth Israel Deaconess Medical Center Boston, Massachusetts Bradley D. Lewis, MD Associate Professor of Radiology Mayo Clinic College of Medicine Consultant in Radiology Mayo Clinic Rochester, Minnesota

viii    Contributors Ana Lourenco, MD Assistant Professor of Diagnostic Imaging Alpert Medical School of Brown University Providence, Rhode Island Edward A. Lyons, OC, FRCPC, FACR Professor of Radiology Obstetrics & Gynecology and Anatomy University of Manitoba Radiologist Health Sciences Center Winnipeg, Manitoba, Canada Giancarlo Mari, MD Professor and Vice-Chair, Department of Obstetrics and Gynecology Director, Division of Maternal-Fetal Medicine University of Tennessee Health Science Center Memphis, Tennessee John R. Mathieson, MD, FRCPC Medical Director and Chief Radiologist Vancouver Island Health Authority Royal Jubilee Hospital Victoria, British Columbia, Canada Cynthia V. Maxwell, MD, FRCSC, RDMS, DABOG Assistant Professor Obstetrics and Gynecology University of Toronto Staff Perinatologist Obstetrics and Gynecology Division of Maternal Fetal Medicine Toronto, Ontario, Canada John McGahan, MD Professor and Vice Chair of Radiology University of California Davis Medical Center Sacramento, California Tejas S. Mehta, MD, MPH Assistant Professor of Radiology Beth Israel Deaconess Medical Center Boston, Massachusetts Christopher R. B. Merritt, BS, MS, MD Professor Thomas Jefferson University Philadelphia, Pennsylvania Norman L. Meyer, MD, PhD Associate Professor, Division of Maternal-Fetal Medicine Vice Chair, Department of OBGYN University of Tennessee Health Science Center Memphis, Tennessee Derek Muradali, MD, FRCPC Head, Division of Ultrasound St. Michael’s Hospital Associate Professor University of Toronto Toronto, Ontario Canada

Sara M. O’Hara, MD, FAAP Associate Professor of Radiology and Pediatrics University of Cincinnati Director, Ultrasound Division Cincinnati Children’s Hospital Medical Center Cincinnati, Ohio †

Heidi B. Patriquin, MD Department of Medical Imaging, Sainte-Justine Hospital Quebec, Canada Joseph F. Polak, MD, MPH Professor of Radiology Tufts University School of Medicine Chief of Radiology Tufts Medical Center Research Affiliation Director, Ultrasound Reading Center Tufts University School of Medicine Boston, Massachusetts Philip Ralls, MD Radiology Professor University of Southern California Keck School of Medicine Los Angeles, California Cynthia T. Rapp, BS, RDMS, FAIUM, FSDMS VP of Clinical Product Development Medipattern Toronto, Ontario, Canada Carl C. Reading, MD, FACR Professor of Radiology Mayo Clinic College of Medicine Consultant in Radiology Mayo Clinic Rochester, Minnesota Maryam Rivaz, MD Post Doctoral Fellow Department of Obstetrics and Gynecology University of Tennessee Health Science Center Memphis, Tennessee Julie E. Robertson, MD, FRCSC Fellow Division of Maternal Fetal Medicine Obstetrics and Gynecology University of Toronto Toronto, Ontario, Canada Henrietta Kotlus Rosenberg, MD, FACR, FAAP Professor of Radiology and Pediatrics The Mount Sinai School of Medicine Director of Pediatric Radiology The Mount Sinai Medical Center New York, New York Carol M. Rumack, MD, FACR Professor of Radiology and Pediatrics University of Colorado Denver School of Medicine Denver, Colorado

†Deceased.

Contributors    ix

Shia Salem, MD, FRCPC Associate Professor University of Toronto Radiologist Mount Sinai Hospital University Health Network Women’s College Hospital Department of Medical Imaging Mount Sinai Hospital Toronto, Ontario, Canada Nathan A. Saucier, MD R4 Resident Diagnostic Radiology University of Illinois College of Medicine at Peoria Peoria, Illinois Eric E. Sauerbrei, BSc, MSc, MD, FRCPC Professor of Radiology, Adjunct Professor of Obstetrics and Gynecology Queen’s University Director of Ultrasound Kingston General Hospital and Hotel Dieu Hospital Director of Residents Research Queen’s University Kingston, Ontario, Canada Joanna J. Seibert, MD Professor of Radiology and Pediatrics Arkansas Children’s Hospital University of Arkansas for Medical Sciences Little Rock, Arkansas Chetan Chandulal Shah, MBBS, DMRD, MBA Assistant Professor Arkansas Children’s Hospital University of Arkansas for Medical Sciences Little Rock, Arkansas Rola Shaheen, MB, BS, MD Radiology Instructor Harvard Medical School Chief of Radiology and Director of Women’s Imaging Harrington Memorial Hospital Boston, Massachusetts William E. Shiels II, DO Chairman, Department of Radiology Nationwide Children’s Hospital Clinical Professor of Radiology, Pediatrics, and Biomedical Engineering The Ohio State University College of Medicine Columbus, Ohio; Adjunct Professor of Radiology The University of Toledo Medical Center Toledo, Ohio Thomas D. Shipp, MD Associate Professor of Obstetrics, Gynecology, and Reproductive Biology Harvard Medical School Boston, Massachusetts Associate Obstetrician and Gynecologist Brigham & Women’s Hospital Boston, Massachusetts

Luigi Solbiati, MD Director, Department of Diagnostic Imaging General Hospital of Busto Arsizio Busto Arsizio, (VA) Italy Elizabeth R. Stamm, MD Associate Professor of Radiology University of Colorado at Denver Health Sciences Center Aurora, Colorado A. Thomas Stavros, MD, FACR Medical Director, Ultrasound Invision Sally Jobe Breast Center Englewood, Colorado George A. Taylor, MD John A. Kirkpatrick Professor of Radiology (Pediatrics) Harvard Medical School Radiologist-in-Chief Children’s Hospital Boston Boston, Massachusetts Wendy Thurston, MD Assistant Professor Department of Medical Imaging University of Toronto Chief, Diagnostic Imaging Department of Diagnostic Imaging St. Joseph’s Health Centre Courtesy Staff Department of Medical Imaging University Health Network Toronto, Ontario, Canada Ants Toi, MD, FRCPC Associate Professor of Radiology and Obstetrics and Gynecology University of Toronto Staff Radiologist University Health Network and Mt. Sinai Hospital Toronto, Ontario, Canada Didier H. Touche, MD Chief Radiologist Centre Sein Godinot Godinot Breast Cancer Center Reims, France Mitchell Tublin, MD Professor of Radiology Chief, Abdominal Imaging Section Department of Radiology University of Pittsburgh School of Medicine Pittsburgh, Pennsylvania Rebecca A. Uhlmann, MS Program Administrator Obstetrics and Gynecology University of Tennessee Health Science Center Memphis, Tennessee Sheila Unger, MD Clinical Geneticist Institute of Human Genetics University of Freiburg Freiburg, Germany

x    Contributors Marnix T. van Holsbeeck, MD Professor of Radiology Wayne State University School of Medicine Detroit, Michigan Division Head, Musculoskeletal Radiology Henry Ford Hospital Detroit, Michigan Patrick M. Vos, MD Clinical Assistant Professor University of British Columbia Vancouver, British Columbia, Canada Dzung Vu, MD, MBBS, Dip Anat Senior Lecturer University of New South Wales Sydney, New South Wales, Australia Wendy L. Whittle, MD Maternal Fetal Medicine Specialist Department of Obstetrics and Gynecology Mount Sinai Hospital University of Toronto Toronto, Ontario, Canada

Stephanie R. Wilson, MD, FRCPC Clinical Professor of Radiology University of Calgary Staff Radiologist Foothills Medical Centre Calgary, Alberta, Canada Rory Windrim, MD, MSc, FRCSC Professor Department of Obstetrics & Gynecology University of Toronto Staff Perinatologist Mount Sinai Hospital Toronto, Ontario, Canada Cynthia E. Withers Staff Radiologist Department of Radiology Santa Barbara Cottage Hospital Santa Barbara, California

In memory of my parents, Drs. Ruth and Raymond Masters, who encouraged me to enjoy the intellectual challenge of medicine and the love of making a difference in patients’ lives.   CMR To a lifetime of clinical colleagues, residents, and fellows who have provided me with a wealth of professional joy. And to my wonderful family, for your love and never-ending support.   SRW To Cathy, Nicholas, Ben, and Laurie, for all the love and joy you bring to my life. You are all I could ever hope for.   JWC To Alex, Becky, and Julie—your love and support made this work possible.

  DL

Preface The fourth edition of Diagnostic Ultrasound is a major revision. Previous editions have been very well accepted as a reference textbook and have been the most commonly used reference in ultrasound education and practices worldwide. We are pleased to provide a new update of images and text with new areas of strength. For the first time we are including video clips in the majority of chapters. The display of real-time ultrasound has helped to capture those abnormalities that require a sweep through the pathology to truly appreciate the lesion. It is similar to scrolling through images on a PACS and has added great value to clinical imaging. Daily we find that cine or video clips show important areas between still images that help to make certain a diagnosis or relationships between lesions. Now we rarely need to go back to reevaluate a lesion with another scan, making patient imaging more efficient. We are pleased to announce that a new editor, Deborah Levine, has joined us, providing expertise in fetal imaging, in both obstetrical sonography and fetal MRI. Prenatal diagnosis is one of the frontiers of medicine that continues to grow as a field and has pushed our understanding of what happens to the fetus before we see a lesion at birth. These antecedents of disease in children and adults help us to arrange care for patients long before the mother goes into labor. Approximately 90 outstanding new and continuing authors have contributed to this edition, and all are recognized experts in the field of ultrasound. We have replaced at least 50% of the images without increasing the size of the two volumes, so new value has been added to all of the chapters, particularly for obstetrics and gynecology. The fourth edition now includes over 5000 images, many in full color. The layout has been exhaustively revamped, and there are highly valuable multipart figures or key figure collages. These images all reflect the spectrum of sonographic changes that may

occur in a given disease instead of the most common manifestation only. The book’s format has been redesigned to facilitate reading and review. There are again color-enhanced boxes to highlight the important or critical features of sonographic diagnoses. Key terms and concepts are emphasized in boldface type. To direct the readers to other research and literature of interest, comprehensive reference lists are organized by topic. Diagnostic Ultrasound is again divided into two volumes. Volume I consists of Parts I to III. Part I contains chapters on physics and biologic effects of ultrasound, as well as more of the latest developments in ultrasound contrast agents. Part II covers abdominal, pelvic, and thoracic sonography, including interventional procedures and organ transplantation. Part III presents small parts imaging including thyroid, breast, scrotum, carotid, peripheral vessels, and particularly MSK imaging. Newly added is a chapter on musculoskeletal intervention. Volume II begins with Part IV, where the greatest expansion of text and images has been on obstetric and fetal sonography, including video clips for the first time. Part V comprehensively covers pediatric sonography. Diagnostic Ultrasound is for practicing physicians, residents, medical students, sonographers, and others interested in understanding the vast applications of diagnostic sonography in patient care. Our goal is for Diagnostic Ultrasound to continue to be the most comprehensive reference book available in the sonographic literature with a highly readable style and superb images. Carol M. Rumack Stephanie R. Wilson J. William Charboneau Deborah Levine

Acknowledgments Our deepest appreciation and sincerest gratitude: To all of our outstanding authors who have contributed extensive, newly updated, and authoritative text and images. We cannot thank them enough for their efforts on this project. To Sharon Emmerling in Denver, Colorado, whose outstanding secretarial and communication skills with authors and editors have facilitated the review and final revision of the entire manuscript. Her enthusiastic attention to detail and accuracy has made this our best edition ever. To Gordana Popovich and Dr. Hojun Yu for their artwork and schematics in Chapter 8, The Gastrointestinal Tract.

To Dr. Hojun Yu for his schematics on liver anatomy in Chapter 4, The Liver. To Lisa Barnes, developmental editor at Elsevier, who has worked closely with us on this project from the very beginning of the fourth edition. We also thank the enthusiastic participation of many other Elsevier experts including Rebecca Gaertner, Elsevier’s guiding hand overseeing the project. She has patiently worked with us through all the final stages of development and production. It has been an intense year for everyone, and we are very proud of this superb edition of Diagnostic Ultrasound.

CHAPTER 1 

Physics of Ultrasound Christopher R. B. Merritt

Chapter Outline BASIC ACOUSTICS Wavelength and Frequency Propagation of Sound Distance Measurement Acoustic Impedance Reflection Refraction Attenuation INSTRUMENTATION Transmitter Transducer Receiver Image Display Mechanical Sector Scanners Arrays Linear Arrays Curved Arrays

A

Phased Arrays Two-Dimensional Arrays

Transducer Selection IMAGE DISPLAY AND STORAGE SPECIAL IMAGING MODES Tissue Harmonic Imaging Spatial Compounding Three-Dimensional Ultrasound IMAGE QUALITY Spatial Resolution IMAGING PITFALLS Shadowing and Enhancement DOPPLER SONOGRAPHY Doppler Signal Processing and Display Doppler Instrumentation Power Mode Doppler

ll diagnostic ultrasound applications are based on the detection and display of acoustic energy reflected from interfaces within the body. These interactions provide the information needed to generate high-resolution, gray-scale images of the body, as well as display information related to blood flow. Its unique imaging attributes have made ultrasound an important and versatile medical imaging tool. However, expensive stateof-the-art instrumentation does not guarantee the production of high-quality studies of diagnostic value. Gaining maximum benefit from this complex technology requires a combination of skills, including knowledge of the physical principles that empower ultrasound with its unique diagnostic capabilities. The user must understand the fundamentals of the interactions of acoustic energy with tissue and the methods and instruments used to produce and optimize the ultrasound display. With this knowledge the user can collect the maximum information from each examination, avoiding pitfalls and errors in diagnosis that may result from the omission of information or the misinterpretation of artifacts. Ultrasound imaging and Doppler ultrasound are based on the scattering of sound energy by interfaces of materials with different properties through interactions governed by acoustic physics. The amplitude of reflected 2

Interpretation of the Doppler Spectrum Interpretation of Color Doppler Other Technical Considerations Doppler Frequency Wall Filters Spectral Broadening Aliasing Doppler Angle Sample Volume Size Doppler Gain

OPERATING MODES: CLINICAL IMPLICATIONS Bioeffects and User Concerns THERAPEUTIC APPLICATIONS: HIGH-INTENSITY FOCUSED ULTRASOUND

energy is used to generate ultrasound images, and frequency shifts in the backscattered ultrasound provide information relating to moving targets such as blood. To produce, detect, and process ultrasound data, users must manage numerous variables, many under their direct control. To do this, operators must understand the methods used to generate ultrasound data and the theory and operation of the instruments that detect, display, and store the acoustic information generated in clinical examinations. This chapter provides an overview of the fundamentals of acoustics, the physics of ultrasound imaging and flow detection, and ultrasound instrumentation with emphasis on points most relevant to clinical practice. A discussion of the therapeutic application of highintensity focused ultrasound concludes the chapter.

BASIC ACOUSTICS Wavelength and Frequency Sound is the result of mechanical energy traveling through matter as a wave producing alternating compression and rarefaction. Pressure waves are propagated by limited physical displacement of the material through

Chapter 1  ■  Physics of Ultrasound   3

FIGURE 1-1.  Sound waves. Sound is transmitted mechanically at the molecular level. In the resting state the pressure is uniform throughout the medium. Sound is propagated as a series of alternating pressure waves producing compression and rarefaction of the conducting medium. The time for a pressure wave to pass a given point is the period, T. The frequency of the wave is 1/T. The wavelength, λ, is the distance between corresponding points on the time-pressure curve.

which the sound is being transmitted. A plot of these changes in pressure is a sinusoidal waveform (Fig. 1-1), in which the Y axis indicates the pressure at a given point and the X axis indicates time. Changes in pressure with time define the basic units of measurement for sound. The distance between corresponding points on the timepressure curve is defined as the wavelength (λ), and the time (T) to complete a single cycle is called the period. The number of complete cycles in a unit of time is the frequency (f ) of the sound. Frequency and period are inversely related. If the period (T) is expressed in seconds, f = 1/T, or f = T × sec–1. The unit of acoustic frequency is the hertz (Hz); 1 Hz = 1 cycle per second. High frequencies are expressed in kilohertz (kHz; 1 kHz = 1000 Hz) or megahertz (MHz; 1 MHz = 1,000,000 Hz). In nature, acoustic frequencies span a range from less than 1 Hz to more than 100,000 Hz (100 kHz). Human hearing is limited to the lower part of this range, extending from 20 to 20,000 Hz. Ultrasound differs from audible sound only in its frequency, and it is 500 to 1000 times higher than the sound we normally hear. Sound frequencies used for diagnostic applications typically range from 2 to 15 MHz, although frequencies as high as 50 to 60 MHz are under investigation for certain specialized imaging applications. In general, the frequencies used for ultrasound imaging are higher than those used for Doppler. Regardless of the frequency, the same basic principles of acoustics apply.

Propagation of Sound In most clinical applications of ultrasound, brief bursts or pulses of energy are transmitted into the body and propagated through tissue. Acoustic pressure waves can travel in a direction perpendicular to the direction of the particles being displaced (transverse waves), but in tissue and fluids, sound propagation is along the direction of particle movement (longitudinal waves). The speed at which the pressure wave moves through tissue varies greatly and is affected by the physical properties of the tissue. Propagation velocity is largely determined by the

Air

330

Fat

1450

Water

1480

Soft tissue (average)

1540

Liver

1550

Kidney

1560

Blood

1570

Muscle

1580

Bone

4080

1400

1500

1600

1700

1800

Propagation velocity (meters/second)

FIGURE 1-2.  Propagation velocity. In the body, propagation velocity of sound is determined by the physical properties of tissue. As shown, this varies considerably. Medical ultrasound devices base their measurements on an assumed average propagation velocity of 1540 m/sec.

resistance of the medium to compression, which in turn is influenced by the density of the medium and its stiffness or elasticity. Propagation velocity is increased by increasing stiffness and reduced by decreasing density. In the body, propagation velocity may be regarded as constant for a given tissue and is not affected by the frequency or wavelength of the sound. Figure 1-2 shows typical propagation velocities for a variety of materials. In the body the propagation velocity of sound is assumed to be 1540 meters per second (m/ sec). This value is the average of measurements obtained from normal tissues.1,2 Although this value represents most soft tissues, such tissues as aerated lung and fat have propagation velocities significantly less than 1540 m/sec, whereas tissues such as bone have greater velocities. Because a few normal tissues have propagation values significantly different from the average value assumed by the ultrasound scanner, the display of such tissues may

4   PART I  ■  Physics

be subject to measurement errors or artifacts (Fig. 1-3). The propagation velocity of sound (c) is related to frequency and wavelength by the following simple equation: c= fλ  1 Thus a frequency of 5 MHz can be shown to have a wavelength of 0.308 mm in tissue: λ = c/f = 1540 m/sec × 5,000,000 sec–1 = 0.000308 m = 0.308 mm.

FIGURE 1-3.  Propagation velocity artifact. When sound passes through a lesion containing fat, echo return is delayed because fat has a propagation velocity of 1450 m/sec, which is less than the liver. Because the ultrasound scanner assumes that sound is being propagated at the average velocity of 1540 m/sec, the delay in echo return is interpreted as indicating a deeper target. Therefore the final image shows a misregistration artifact in which the diaphragm and other structures deep to the fatty lesion are shown in a deeper position than expected (simulated image).

FIGURE 1-4.  Ultrasound ranging. The information used to position an echo for display is based on the precise measurement of time. Here the time for an echo to travel from the transducer to the target and return to the transducer is 0.145 ms. Multiplying the velocity of sound in tissue (1540 m/sec) by the time shows that the sound returning from the target has traveled 22.33 cm. Therefore the target lies half this distance, or 11.165 cm, from the transducer.

Distance Measurement Propagation velocity is a particularly important value in clinical ultrasound and is critical in determining the distance of a reflecting interface from the transducer. Much of the information used to generate an ultrasound scan is based on the precise measurement of time and employs the principles of echo-ranging. If an ultrasound pulse is transmitted into the body and the time until an echo returns is measured, it is simple to calculate the depth of the interface that generated the echo, provided the propagation velocity of sound for the tissue is known. For example, if the time from the transmission of a pulse until the return of an echo is 0.145 millisecond (ms; 0.000145 sec) and the velocity of sound is 1540 m/sec, the distance that the sound has traveled must be 22.33 cm (1540 m/sec × 100 cm/m × 0.000145 sec = 22.33 cm). Because the time measured includes the time for sound to travel to the interface and then return along the same path to the transducer, the distance from the transducer to the reflecting interface is 22.33 cm/2 = 11.165 cm (Fig. 1-4). The accuracy of this measurement is therefore highly influenced by how closely the presumed velocity of sound corresponds to the true velocity in the tissue being observed (see Figs. 1-2 and 1-3), as well as by the important assumption that the sound pulse travels in a straight path to and from the reflecting interface.

Acoustic Impedance Current diagnostic ultrasound scanners rely on the detection and display of reflected sound or echoes. Imaging based on transmission of ultrasound is also possible, but this is not used clinically at present. To produce an echo, a reflecting interface must be present. Sound passing through a totally homogeneous medium

Chapter 1  ■  Physics of Ultrasound   5

encounters no interfaces to reflect sound, and the medium appears anechoic or cystic. At the junction of tissues or materials with different physical properties, acoustic interfaces are present. These interfaces are responsible for the reflection of variable amounts of the incident sound energy. Thus, when ultrasound passes from one tissue to another or encounters a vessel wall or circulating blood cells, some of the incident sound energy is reflected. The amount of reflection or backscatter is determined by the difference in the acoustic impedances of the materials forming the interface. Acoustic impedance (Z) is determined by product of the density (ρ) of the medium propagating the sound and the propagation velocity (c) of sound in that medium (Z = ρc). Interfaces with large acoustic impedance differences, such as interfaces of tissue with air or bone, reflect almost all the incident energy. Interfaces composed of substances with smaller differences in acoustic impedance, such as a muscle and fat interface, reflect only part of the incident energy, permitting the remainder to continue onward. As with propagation velocity, acoustic impedance is determined by the properties of the tissues involved and is independent of frequency.

of energy reflected by an acoustic interface can be expressed as a fraction of the incident energy; this is termed the reflection coefficient (R). If a specular reflector is perpendicular to the incident sound beam, the amount of energy reflected is determined by the following relationship: R = (Z 2 − Z1 )2 (Z 2 + Z1 )2   2 where Z1 and Z2 are the acoustic impedances of the media forming the interface. Because ultrasound scanners only detect reflections that return to the transducer, display of specular interfaces is highly dependent on the angle of insonation (exposure to ultrasound waves). Specular reflectors will return echoes to the transducer only if the sound beam is perpendicular to the interface. If the interface is not at a 90-degree angle to the sound beam, it will be reflected away from the transducer, and the echo will not be detected (see Fig. 1-5, A). Most echoes in the body do not arise from specular reflectors but rather from much smaller interfaces within solid organs. In this case the acoustic interfaces involve structures with individual dimensions much smaller than

Reflection EXAMPLES OF SPECULAR REFLECTORS

The way ultrasound is reflected when it strikes an acoustic interface is determined by the size and surface features of the interface (Fig. 1-5). If large and relatively smooth, the interface reflects sound much as a mirror reflects light. Such interfaces are called specular reflectors because they behave as “mirrors for sound.” The amount

A

Diaphragm Wall of urine-filled bladder Endometrial stripe

B

FIGURE 1-5.  Specular and diffuse reflectors. A, Specular reflector. The diaphragm is a large and relatively smooth surface that reflects sound like a mirror reflects light. Thus, sound striking the diaphragm at nearly a 90-degree angle is reflected directly back to the transducer, resulting in a strong echo. Sound striking the diaphragm obliquely is reflected away from the transducer, and an echo is not displayed (yellow arrow). B, Diffuse reflector. In contrast to the diaphragm, the liver parenchyma consists of acoustic interfaces that are small compared to the wavelength of sound used for imaging. These interfaces scatter sound in all directions, and only a portion of the energy returns to the transducer to produce the image.

6   PART I  ■  Physics θ1 = 70°

Tissue A C1 = 1540 m/sec Tissue B C2 = 1450 m/sec

θ2 = 86°

FIGURE 1-6.  Ultrasound speckle. Close inspection of an ultrasound image of the breast containing a small cyst reveals it to be composed of numerous areas of varying intensity (speckle). Speckle results from the constructive (red) and destructive (green) interaction of the acoustic fields (yellow rings) generated by the scattering of ultrasound from small tissue reflectors. This interference pattern gives ultrasound images their characteristic grainy appearance and may reduce contrast. Ultrasound speckle is the basis of the texture displayed in ultrasound images of solid tissues.

FIGURE 1-7.  Refraction. When sound passes from tissue A with one acoustic propagation velocity (c1) to tissue B with a different propagation velocity (c2), there is a change in the direction of the sound wave because of refraction. The degree of change is related to the ratio of the propagating velocities of the media forming the interface (sinθ1/sinθ2 = c1/c2).

the wavelength of the incident sound. The echoes from these interfaces are scattered in all directions. Such reflectors are called diffuse reflectors and account for the echoes that form the characteristic echo patterns seen in solid organs and tissues (see Fig. 1-5, B). The constructive and destructive interference of sound scattered by diffuse reflectors results in the production of ultrasound speckle, a feature of tissue texture of sonograms of solid organs (Fig. 1-6). For some diagnostic applications, the nature of the reflecting structures creates important conflicts. For example, most vessel walls behave as specular reflectors that require insonation at a 90-degree angle for best imaging, whereas Doppler imaging requires an angle of less than 90 degrees between the sound beam and the vessel.

Refraction Another event that can occur when sound passes from a tissue with one acoustic propagation velocity to a tissue with a higher or lower sound velocity is a change in the direction of the sound wave. This change in direction of propagation is called refraction and is governed by Snell’s law: sin θ1 sin θ2 = c1 c 2   3 where θ1 is the angle of incidence of the sound approaching the interface, θ2 is the angle of refraction, and c1 and c2 are the propagation velocities of sound in the media

FIGURE 1-8.  Refraction artifact. Axial transabdominal image of the uterus shows a small gestational sac (A) and what appears to be a second sac (B). In this case, the artifact B is caused by refraction at the edge of the rectus abdominis muscle. The bending of the path of the sound results in the creation of a duplicate of the image of the sac in an unexpected and misleading location (simulated image).

forming the interface (Fig. 1-7). Refraction is important because it is one cause of misregistration of a structure in an ultrasound image (Fig. 1-8). When an ultrasound scanner detects an echo, it assumes that the source of the echo is along a fixed line of sight from the transducer. If

Chapter 1  ■  Physics of Ultrasound   7

the sound has been refracted, the echo detected may be coming from a different depth or location than the image shown in the display. If this is suspected, increasing the scan angle so that it is perpendicular to the interface minimizes the artifact.

Attenuation As the acoustic energy moves through a uniform medium, work is performed and energy is ultimately transferred to the transmitting medium as heat. The capacity to perform work is determined by the quantity of acoustic energy produced. Acoustic power, expressed in watts (W) or milliwatts (mW), describes the amount of acoustic energy produced in a unit of time. Although measurement of power provides an indication of the energy as it relates to time, it does not take into account the spatial distribution of the energy. Intensity (I) is used to describe the spatial distribution of power and is calculated by dividing the power by the area over which the power is distributed, as follows: I ( W cm 2 ) = Power ( W ) Area (cm 2 )   4 The attenuation of sound energy as it passes through tissue is of great clinical importance because it influences the depth in tissue, from which useful information can be obtained. This in turn affects transducer selection and a number of operator-controlled instrument settings, including time (or depth) gain compensation, power output attenuation, and system gain levels. Attenuation is measured in relative rather than absolute units. The decibel (dB) notation is generally used to compare different levels of ultrasound power or intensity. This value is 10 times the log10 of the ratio of the power or intensity values being compared. For example, if the intensity measured at one point in tissues is 10 mW/cm2 and at a deeper point is 0.01 mW/cm2, the difference in intensity is as follows: (10) ( log10 0.01 10) = (10) ( log10 0.001) = (10) ( − log10 1000) = (10) ( −3) = −30 dB

As it passes through tissue, sound loses energy, and the pressure waves decrease in amplitude as they travel farther from their source. Contributing to the attenuation of sound are the transfer of energy to tissue, resulting in heating (absorption), and the removal of energy by reflection and scattering. Attenuation is therefore the result of the combined effects of absorption, scattering, and reflection. Attenuation depends on the insonating frequency as well as the nature of the attenuating medium. High frequencies are attenuated more rapidly than lower frequencies, and transducer frequency is a major determinant of the useful depth from which information can be obtained with ultrasound. Attenuation determines the efficiency with which ultrasound penetrates a specific tissue and varies considerably in normal tissues (Fig. 1-9).

Water

0.00

Blood

0.18

Fat

0.63

Soft tissue (average)

0.70

Liver

0.94

Kidney

1.00

Muscle (parallel) Muscle (transverse)

1.30 3.30

Bone

5.00

Air

10.00 0

2

4

6

8

10

Attenuation (dB/cm/MHz)

FIGURE 1-9.  Attenuation. As sound passes through tissue, it loses energy through the transfer of energy to tissue by heating, reflection, and scattering. Attenuation is determined by the insonating frequency and the nature of the attenuating medium. Attenuation values for normal tissues show considerable variation. Attenuation also increases in proportion to insonating frequency, resulting in less penetration at higher frequencies.

INSTRUMENTATION Ultrasound scanners are complex and sophisticated imaging devices, but all consist of the following basic components to perform key functions: • Transmitter or pulser to energize the transducer • Ultrasound transducer itself • Receiver and processor to detect and amplify the backscattered energy and manipulate the reflected signals for display • Display that presents the ultrasound image or data in a form suitable for analysis and interpretation • Method to record or store the ultrasound image

Transmitter Most clinical applications use pulsed ultrasound, in which brief bursts of acoustic energy are transmitted into the body. The source of these pulses, the ultrasound transducer, is energized by application of precisely timed, high-amplitude voltage. The maximum voltage that may be applied to the transducer is limited by federal regulations that restrict the acoustic output of diagnostic scanners. Most scanners provide a control that permits attenuation of the output voltage. Because the use of maximum output results in higher exposure of the patient to ultrasound energy, prudent use dictates use of the output attenuation controls to reduce power levels to the lowest levels consistent with the diagnostic problem.3

8   PART I  ■  Physics

The transmitter also controls the rate of pulses emitted by the transducer, or the pulse repetition frequency (PRF). The PRF determines the time interval between ultrasound pulses and is important in determining the depth from which unambiguous data can be obtained both in imaging and Doppler modes. The ultrasound pulses must be spaced with enough time between the pulses to permit the sound to travel to the depth of interest and return before the next pulse is sent. For imaging, PRFs from 1 to 10 kHz are used, resulting in an interval of 0.1 to 1 ms between pulses. Thus, a PRF of 5 kHz permits an echo to travel and return from a depth of 15.4 cm before the next pulse is sent.

Transducer A transducer is any device that converts one form of energy to another. In ultrasound the transducer converts electric energy to mechanical energy, and vice versa. In diagnostic ultrasound systems the transducer serves two functions: (1) converting the electric energy provided by the transmitter to the acoustic pulses directed into the patient and (2) serving as the receiver of reflected echoes, converting weak pressure changes into electric signals for processing. Ultrasound transducers use piezoelectricity, a principle discovered by Pierre and Jacques Curie in 1880. Piezoelectric materials have the unique ability to respond to the action of an electric field by changing shape. They also have the property of generating electric potentials when compressed. Changing the polarity of a voltage applied to the transducer changes the thickness of the transducer, which expands and contracts as the polarity changes. This results in the generation of mechanical pressure waves that can be transmitted into the body. The piezoelectric effect also results in the generation of small potentials across the transducer when the transducer is struck by returning echoes. Positive pressures cause a small polarity to develop across the transducer; negative pressure during the rarefaction portion of the acoustic wave produces the opposite polarity across the transducer. These tiny polarity changes and the associated voltages are the source of all the information processed to generate an ultrasound image or Doppler display. When stimulated by the application of a voltage difference across its thickness, the transducer vibrates. The frequency of vibration is determined by the transducer material. When the transducer is electrically stimulated, a range or band of frequencies results. The preferential frequency produced by a transducer is determined by the propagation speed of the transducer material and its thickness. In the pulsed wave operating modes used for most clinical ultrasound applications, the ultrasound pulses contain additional frequencies that are both higher and lower than the preferential frequency. The range of frequencies produced by a given transducer is termed its

bandwidth. Generally, the shorter the pulse of ultrasound produced by the transducer, the greater is the bandwidth. Most modern digital ultrasound systems employ broad-bandwidth technology. Ultrasound bandwidth refers to the range of frequencies produced and detected by the ultrasound system. This is important because each tissue in the body has a characteristic response to ultrasound of a given frequency, and different tissues respond differently to different frequencies. The range of frequencies arising from a tissue exposed to ultrasound is referred to as the frequency spectrum bandwidth of the tissue, or tissue signature. Broad-bandwidth technology provides a means to capture the frequency spectrum of insonated tissues, preserving acoustic information and tissue signature. Broad-bandwidth beam formers reduce speckle artifact by a process of frequency compounding. This is possible because speckle patterns at different frequencies are independent of one another, and combining data from multiple frequency bands (i.e., compounding) results in a reduction of speckle in the final image, leading to improved contrast resolution. The length of an ultrasound pulse is determined by the number of alternating voltage changes applied to the transducer. For continuous wave (CW) ultrasound devices, a constant alternating current is applied to the transducer, and the alternating polarity produces a continuous ultrasound wave. For imaging, a single, brief voltage change is applied to the transducer, causing it to vibrate at its preferential frequency. Because the transducer continues to vibrate or “ring” for a short time after it is stimulated by the voltage change, the ultrasound pulse will be several cycles long. The number of cycles of sound in each pulse determines the pulse length. For imaging, short pulse lengths are desirable because longer pulses result in poorer axial resolution. To reduce the pulse length to no more than two or three cycles, damping materials are used in the construction of the transducer. In clinical imaging applications, very short pulses are applied to the transducer, and the transducers have highly efficient damping. This results in very short pulses of ultrasound, generally consisting of only two or three cycles of sound. The ultrasound pulse generated by a transducer must be propagated in tissue to provide clinical information. Special transducer coatings and ultrasound coupling gels are necessary to allow efficient transfer of energy from the transducer to the body. Once in the body, the ultrasound pulses are propagated, reflected, refracted, and absorbed, in accordance with the basic acoustic principles summarized earlier. The ultrasound pulses produced by the transducer result in a series of wavefronts that form a three-dimensional (3-D) beam of ultrasound. The features of this beam are influenced by constructive and destructive interference of the pressure waves, the curvature of the transducer, and acoustic lenses used to shape the beam.

Chapter 1  ■  Physics of Ultrasound   9

Interference of pressure waves results in an area near the transducer where the pressure amplitude varies greatly. This region is termed the near field, or Fresnel zone. Farther from the transducer, at a distance determined by the radius of the transducer and the frequency, the sound field begins to diverge, and the pressure amplitude decreases at a steady rate with increasing distance from the transducer. This region is called the far field, or Fraunhofer zone. In modern multielement transducer arrays, precise timing of the firing of elements allows correction of this divergence of the ultrasound beam and focusing at selected depths. Only reflections of pulses that return to the transducer are capable of stimulating the transducer with small pressure changes, which are converted into the voltage changes that are detected, amplified, and processed to build an image based on the echo information.

Receiver When returning echoes strike the transducer face, minute voltages are produced across the piezoelectric elements. The receiver detects and amplifies these weak signals. The receiver also provides a means for compensating for the differences in echo strength, which result from attenuation by different tissue thickness by control of time gain compensation (TGC) or depth gain compensation (DGC). Sound is attenuated as it passes into the body, and additional energy is removed as echoes return through tissue to the transducer. The attenuation of sound is proportional to the frequency and is constant for specific tissues. Because echoes returning from deeper tissues are weaker than those returning from more superficial structures, they must be amplified more by the receiver to produce a uniform tissue echo appearance (Fig. 1-10). This adjustment is accomplished by TGC controls that permit the user to selectively amplify the signals from deeper structures or to suppress the signals from super-

ficial tissues, compensating for tissue attenuation. Although many newer machines provide for some means of automatic TGC, the manual adjustment of this control is one of the most important user controls and may have a profound effect on the quality of the ultrasound image provided for interpretation. Another important function of the receiver is the compression of the wide range of amplitudes returning to the transducer into a range that can be displayed to the user. The ratio of the highest to the lowest amplitudes that can be displayed may be expressed in decibels and is referred to as the dynamic range. In a typical clinical application, the range of reflected signals may vary by a factor of as much as 1 : 1012, resulting in a dynamic range of up to 120 dB. Although the amplifiers used in ultrasound machines are capable of handling this range of voltages, gray-scale displays are limited to display a signal intensity range of only 35 to 40 dB. Compression and remapping of the data are required to adapt the dynamic range of the backscattered signal intensity to the dynamic range of the display (Fig. 1-11). Compression is performed in the receiver by selective amplification of weaker signals. Additional manual postprocessing controls permit the user to map selectively the returning signal to the display. These controls affect the brightness of different echo levels in the image and therefore determine the image contrast.

Image Display Ultrasound signals may be displayed in several ways.4 Over the years, imaging has evolved from simple A-mode and bistable display to high-resolution, real-time, grayscale imaging. The earliest A-mode devices displayed the voltage produced across the transducer by the backscattered echo as a vertical deflection on the face of an oscilloscope. The horizontal sweep of the oscilloscope was calibrated to indicate the distance from the transducer to the reflecting surface. In this form of display, the

FIGURE 1-10.  Time gain compensation (TGC). Without TGC, tissue attenuation causes gradual loss of display of deeper tissues (A). In this example, tissue attenuation of 1 dB/cm-MHz is simulated for a transducer of 10 MHz. At a depth of 2 cm, the intensity is −20 dB (1% of initial value). By applying increasing amplification or gain to the backscattered signal to compensate for this attenuation, a uniform intensity is restored to the tissue at all depths (B).

10   PART I  ■  Physics

FIGURE 1-11. Dynamic range. The ultrasound receiver must compress the wide range of amplitudes returning to the transducer into a range that can be displayed to the user. Here, compression and remapping of the data to display dynamic ranges of 35, 40, 50, and 60 dB are shown. The widest dynamic range shown (60 dB) permits the best differentiation of subtle differences in echo intensity and is preferred for most imaging applications. The narrower ranges increase conspicuity of larger echo differences.

A

C

B

A

C

B

FIGURE 1-12.  M-mode display. M-mode ultrasound displays changes of echo amplitude and position with time. Display of changes in echo position is useful in the evaluation of rapidly moving structures such as cardiac valves and chamber walls. Here, the three major moving structures in an M-mode image of the fetal heart correspond to the near ventricular wall (A), the interventricular septum (B), and the far ventricular wall (C). The baseline is a time scale that permits the calculation of heart rate from the M-mode data.

strength or amplitude of the reflected sound is indicated by the height of the vertical deflection displayed on the oscilloscope. With A-mode ultrasound, only the position and strength of a reflecting structure are recorded. Another simple form of imaging, M-mode ultrasound, displays echo amplitude and shows the position of moving reflectors (Fig. 1-12). M-mode imaging uses the brightness of the display to indicate the intensity of the reflected signal. The time base of the display can be adjusted to allow for varying degrees of temporal resolution, as dictated by clinical application. M-mode ultrasound is interpreted by assessing motion patterns of specific reflectors and determining anatomic relationships from characteristic patterns of motion. Currently, the major application of M-mode display is evaluation of the rapid motion of cardiac valves and of cardiac

chamber and vessel walls. M-mode imaging may play a future role in measurement of subtle changes in vessel wall elasticity accompanying atherogenesis. The mainstay of imaging with ultrasound is provided by real-time, gray-scale, B-mode display, in which variations in display intensity or brightness are used to indicate reflected signals of differing amplitude. To generate a two-dimensional (2-D) image, multiple ultrasound pulses are sent down a series of successive scan lines (Fig. 1-13), building a 2-D representation of echoes arising from the object being scanned. When an ultrasound image is displayed on a black background, signals of greatest intensity appear as white; absence of signal is shown as black; and signals of intermediate intensity appear as shades of gray. If the ultrasound beam is moved with respect to the object being examined and the position of the reflected signal is stored, the brightest portions of the resulting 2-D image indicate structures reflecting more of the transmitted sound energy back to the transducer. In most modern instruments a digital memory of 512 × 512 or 512 × 640 pixels is used to store values that correspond to the echo intensities originating from corresponding positions in the patient. At least 28, or 256, shades of gray are possible for each pixel, in accord with the amplitude of the echo being represented. The image stored in memory in this manner can then be sent to a video monitor for display. Because B-mode display relates the strength of a backscattered signal to a brightness level on the display device (usually a video display monitor), it is important that the operator understand how the amplitude information in the ultrasound signal is translated into a brightness scale in the image display. Each ultrasound manufacturer offers several options for the way the dynamic range of the target is compressed for display, as well as the transfer function that assigns a given signal amplitude to a shade of gray. Although these technical details vary among machines, the way the operator uses them may greatly affect the clinical value of the final image. In general, it

Chapter 1  ■  Physics of Ultrasound   11

FIGURE 1-13.  B-mode imaging. A 2-D, real-time image is built by ultrasound pulses sent down a series of successive scan lines. Each scan line adds to the image, building a 2-D representation of echoes from the object being scanned. In real-time imaging, an entire image is created 15 to 60 times per second.

is desirable to display as wide a dynamic range as possible, to identify subtle differences in tissue echogenicity (see Fig. 1-11). Real-time ultrasound produces the impression of motion by generating a series of individual 2-D images at rates of 15 to 60 frames per second. Real-time, 2-D, B-mode ultrasound is now the major method for ultrasound imaging throughout the body and is the most common form of B-mode display. Real-time ultrasound permits assessment of both anatomy and motion. When images are acquired and displayed at rates of several times per second, the effect is dynamic, and because the image reflects the state and motion of the organ at the time it is examined, the information is regarded as being shown in real time. In cardiac applications the terms “2-D echocardiography” and “2-D echo” are used to describe real-time, B-mode imaging; in most other applications the term “real-time ultrasound” is used. Transducers used for real-time imaging may be classified by the method used to steer the beam in rapidly generating each individual image, keeping in mind that as many as 30 to 60 complete images must be generated

per second for real-time applications. Beam steering may be done through mechanical rotation or oscillation of the transducer or by electronic means (Fig. 1-14). Electronic beam steering is used in linear array and phased array transducers and permits a variety of image display formats. Most electronically steered transducers currently in use also provide electronic focusing that is adjustable for depth. Mechanical beam steering may use single-element transducers with a fixed focus or may use annular arrays of elements with electronically controlled focusing. For real-time imaging, transducers using mechanical or electronic beam steering generate displays in a rectangular or pie-shaped format. For obstetric, small parts, and peripheral vascular examinations, linear array transducers with a rectangular image format are often used. The rectangular image display has the advantage of a larger field of view near the surface but requires a large surface area for transducer contact. Sector scanners with either mechanical or electronic steering require only a small surface area for contact and are better suited for examinations in which access is limited.

12   PART I  ■  Physics

A

FIGURE 1-14.  Beam steering. A, Linear array. In a linear array transducer, individual elements or groups of elements are fired in sequence. This generates a series of parallel ultrasound beams, each perpendicular to the transducer face. As these beams move across the transducer face, they generate the lines of sight that combine to form the final image. Depending on the number of transducer elements and the sequence in which they are fired, focusing at selected depths from the surface can be achieved. B, Phased array. A phased array transducer produces a sector field of view by firing multiple transducer elements in precise sequence to generate interference of acoustic wavefronts. The ultrasound beam that results generates a series of lines of sight at varying angles from one side of the transducer to the other, producing a sector image format.

B

Mechanical Sector Scanners Early ultrasound scanners used transducers consisting of a single piezoelectric element. To generate real-time images with these transducers, mechanical devices were required to move the transducer in a linear or circular motion. Mechanical sector scanners using one or more single-element transducers do not allow variable focusing. This problem is overcome by using annular array

transducers. Although important in the early days of real-time imaging, mechanical sector scanners with fixed-focus, single-element transducers are not presently in common use.

Arrays Current technology uses a transducer composed of multiple elements, usually produced by precise slicing of a

Chapter 1  ■  Physics of Ultrasound   13

piece of piezoelectric material into numerous small units, each with its own electrodes. Such transducer arrays may be formed in a variety of configurations. Typically, these are linear, curved, phased, or annular arrays. High-density 2-D arrays have also been developed. By precise timing of the firing of combinations of elements in these arrays, interference of the wavefronts generated by the individual elements can be exploited to change the direction of the ultrasound beam, and this can be used to provide a steerable beam for the generation of real-time images in a linear or sector format.

Linear Arrays Linear array transducers are used for small parts, vascular, and obstetric applications because the rectangular image format produced by these transducers is well suited for these applications. In these transducers, individual elements are arranged in a linear fashion. By firing the transducer elements in sequence, either individually or in groups, a series of parallel pulses is generated, each forming a line of sight perpendicular to the transducer face. These individual lines of sight combine to form the image field of view (see Fig. 1-14, A). Depending on the number of transducer elements and the sequence in which they are fired, focusing at selected depths from the surface can be achieved.

Curved Arrays Linear arrays that have been shaped into convex curves produce an image that combines a relatively large surface field of view with a sector display format. Curved array transducers are used for a variety of applications, the larger versions serving for general abdominal, obstetric, and transabdominal pelvic scanning. Small, high-frequency, curved array scanners are often used in transvaginal and transrectal probes and for pediatric imaging.

Phased Arrays In contrast to mechanical sector scanners, phased array scanners have no moving parts. A sector field of view is produced by multiple transducer elements fired in precise sequence under electronic control. By controlling the time and sequence at which the individual transducer elements are fired, the resulting ultrasound wave can be steered in different directions as well as focused at different depths (see Fig. 1-14, B). By rapidly steering the beam to generate a series of lines of sight at varying angles from one side of the transducer to the other, a sector image format is produced. This allows the fabrication of transducers of relatively small size but with large fields of view at depth. These transducers are particularly useful for intercostal scanning, to evaluate the heart, liver, or spleen, and for examinations in other areas where access is limited.

FIGURE 1-15.  Two-dimensional array. High-density 2-D arrays consist of a 2-D matrix of transducer elements, permitting acquisition of data from a volume rather than a single plane of tissue. Precise electronic control of individual elements permits adjustable focusing on both azimuth and elevation planes.

Two-Dimensional Arrays Transducer arrays can be formed (1) by slicing a rectangular piece of transducer material perpendicular to its long axis to produce a number of small rectangular elements or (2) by creating a series of concentric elements nested within one another in a circular piece of piezoelectric material to produce an annular array. The use of multiple elements permits precise focusing. A particular advantage of 2-D array construction is that the beam can be focused in both the elevation plane and the lateral plane, and a uniform and highly focused beam can be produced (Fig. 1-15). These arrays improve spatial resolution and contrast, reduce clutter, and are well suited for the collection of data from volumes of tissue for use in 3-D processing and display. Unlike linear 2-D arrays, in which delays in the firing of the individual elements may be used to steer the beam, annular arrays do not permit beam steering and, to be used for real-time imaging, must be steered mechanically.

Transducer Selection Practical considerations in the selection of the optimal transducer for a given application include not only the requirements for spatial resolution, but also the distance of the target object from the transducer because penetration of ultrasound diminishes as frequency increases. In general, the highest ultrasound frequency permitting penetration to the depth of interest should be selected. For superficial vessels and organs, such as the thyroid, breast, or testicle, lying within 1 to 3 cm of the surface, imaging frequencies of 7.5 to 15 MHz are typically used. These high frequencies are also ideal for intraoperative

14   PART I  ■  Physics

applications. For evaluation of deeper structures in the abdomen or pelvis more than 12 to 15 cm from the surface, frequencies as low as 2.25 to 3.5 MHz may be required. When maximal resolution is needed, a highfrequency transducer with excellent lateral and elevation resolution at the depth of interest is required.

IMAGE DISPLAY AND STORAGE With real-time ultrasound, user feedback is immediate and is provided by video display. The brightness and contrast of the image on this display are determined by the ambient lighting in the examination room, the brightness and contrast settings of the video monitor, the system gain setting, and the TGC adjustment. The factor most affecting image quality in many ultrasound departments is probably improper adjustment of the video display, with a lack of appreciation of the relationship between the video display settings and the appearance of hard copy or images viewed on a workstation. Because of the importance of the real-time video display in providing feedback to the user, it is essential that the display and the lighting conditions under which it is viewed are standardized and matched to the display used for interpretation. Interpretation of images and archival storage of images may be in the form of transparencies printed on film by optical or laser cameras and printers, videotape, or digital picture archiving and communications system (PACS). Increasingly, digital storage is being used for archiving of ultrasound images.

SPECIAL IMAGING MODES Tissue Harmonic Imaging Variation of the propagation velocity of sound in fat and other tissues near the transducer results in a phase aberration that distorts the ultrasound field, producing noise and clutter in the ultrasound image. Tissue harmonic imaging provides an approach for reducing the effects of phase aberrations.5 Nonlinear propagation of ultrasound through tissue is associated with the more rapid propagation of the high-pressure component of the ultrasound pressure wave than its negative (rarefactional) component. This results in increasing distortion of the acoustic pulse as it travels within the tissue and causes the generation of multiples, or harmonics, of the transmitted frequency (Fig. 1-16). Tissue harmonic imaging takes advantage of the generation, at depth, of these harmonics. Because the generation of harmonics requires interaction of the transmitted field with the propagating tissue, harmonic generation is not present near the transducer/skin interface, and it only becomes important some distance from the transducer. In most cases the near and far fields of

FIGURE 1-16.  Harmonic generation. The transmitted waveform is shown in A. As the sound is propagated through tissue, the high-pressure component of the wave travels more rapidly than the rarefactional component, producing distortion (B) of the wave and generating higher-frequency components (harmonics). (From Merritt CR: Technology update. Radiol Clin North Am 2001;39:385-397.)

the image are affected less by harmonics than by intermediate locations. Using broad-bandwidth transducers and signal filtration or coded pulses, the harmonic signals reflected from tissue interfaces can be selectively displayed. Because most imaging artifacts are caused by the interaction of the ultrasound beam with superficial structures or by aberrations at the edges of the beam profile, these artifacts are eliminated using harmonic imaging because the artifact-producing signals do not consist of sufficient energy to generate harmonic frequencies and therefore are filtered out during image formation. Images generated using tissue harmonics often exhibit reduced noise and clutter (Fig. 1-17). Because harmonic beams are narrower than the originally transmitted beams, spatial resolution is improved and side lobes are reduced.

Spatial Compounding An important source of image degradation and loss of contrast is ultrasound speckle. Speckle results from the constructive and destructive interaction of the acoustic fields generated by the scattering of ultrasound from small tissue reflectors. This interference pattern gives ultrasound images their characteristic grainy appearance (see Fig. 1-6), reducing contrast (Fig. 1-18) and making the identification of subtle features more difficult. By summing images from different scanning angles through compound scanning (Fig. 1-19), significant improvement in the contrast-to-noise (speckle) ratio can be achieved (Fig. 1-20). This is because speckle is random, and the generation of an image by compounding will reduce speckle noise because only the signal is reinforced. In addition, spatial compounding may reduce artifacts that result when an ultrasound beam strikes a specular

Chapter 1  ■  Physics of Ultrasound   15

FIGURE 1-17.  Tissue harmonic imaging. A, Conventional image, and B, tissue harmonic image, of gallbladder of patient with acute cholecystitis. Note the reduction of noise and clutter in the tissue harmonic image. Because harmonic beams do not interact with superficial structures and are narrower than the originally transmitted beam, spatial resolution is improved and clutter and side lobes are reduced. (From Merritt CR: Technology update. Radiol Clin North Am 2001;39:385-397.)

A

A

B

B

FIGURE 1-18.  Effect of speckle on contrast. A, Speckle noise partially obscures the simulated lesion. B, The speckle has been reduced, increasing contrast resolution between the lesion and the background. (From Merritt CR: Technology update. Radiol Clin North Am 2001;39:385-397.)

FIGURE 1-19.  Spatial compounding. A, Conventional imaging is limited to a fixed angle of incidence of ultrasound scan lines to tissue interfaces, resulting in poor definition of specular reflectors that are not perpendicular to the beam. B, Spatial compounding combines images obtained by insonating the target from multiple angles. In addition to improving detection interfaces, compounding reduces speckle noise because only the signal is reinforced; speckle is random and not reinforced. This improves contrast.

16   PART I  ■  Physics

FIGURE 1-20.  Spatial compounding. A, Conventional image, and B, compound image, of the thyroid. Note the reduced speckle as well as better definition of superficial tissue (blue arrow) as well as small cysts (yellow arrows) and calcifications (white arrow).

A

reflector at an angle greater or less than 90 degrees. In conventional real-time imaging, each scan line used to generate the image strikes the target at a constant, fixed angle. As a result, strong reflectors that are not perpendicular to the ultrasound beam scatter sound in directions that prevent their clear detection and display. This in turn results in poor margin definition and less distinct boundaries for cysts and other masses. Compounding has been found to reduce these artifacts. Limitations of compounding are diminished visibility of shadowing and enhancement; however, these are offset by the ability to evaluate lesions, both with and without compounding, preserving shadowing and enhancement when these features are important to diagnosis.6

Three-Dimensional Ultrasound Dedicated 3-D scanners used for fetal, gynecologic, and cardiac scanning may employ hardware-based image registration, high-density 2-D arrays, or software registration of scan planes as a tissue volume is acquired. 3-D

B

imaging permits volume data to be viewed in multiple imaging planes and allows accurate measurement of lesion volume (Fig. 1-21).

IMAGE QUALITY The key determinants of the quality of an ultrasound image are its spatial, contrast, and temporal resolution, as well as freedom from certain artifacts.

Spatial Resolution The ability to differentiate two closely situated objects as distinct structures is determined by the spatial resolution of the ultrasound device. Spatial resolution must be considered in three planes, with different determinants of resolution for each. Simplest is the resolution along the axis of the ultrasound beam, or axial resolution. With pulsed wave ultrasound, the transducer introduces a series of brief bursts of sound into the body. Each

Chapter 1  ■  Physics of Ultrasound   17

slices of information from the body, and the width and thickness of the ultrasound beam are important determinants of image quality. Excessive beam width and thickness limit the ability to delineate small features and may obscure shadowing and enhancement from small structures, such as breast microcalcifications and small thyroid cysts. The width and thickness of the ultrasound beam determine lateral resolution and elevation resolution, respectively. Lateral and elevation resolutions are significantly poorer than the axial resolution of the beam. Lateral resolution is controlled by focusing the beam, usually by electronic phasing, to alter the beam width at a selected depth of interest. Elevation resolution is determined by the construction of the transducer and generally cannot be controlled by the user.

IMAGING PITFALLS

FIGURE 1-21.  3-D ultrasound image, 24-week fetus. Three-dimensional ultrasound permits collection and review of data obtained from a volume of tissue in multiple imaging planes, as well as a rendering of surface features.

ultrasound pulse typically consists of two or three cycles of sound. The pulse length is the product of the wavelength and the number of cycles in the pulse. Axial resolution, the maximum resolution along the beam axis, is determined by the pulse length (Fig. 1-22). Because ultrasound frequency and wavelength are inversely related, the pulse length decreases as the imaging frequency increases. Because the pulse length determines the maximum resolution along the axis of the ultrasound beam, higher transducer frequencies provide higher image resolution. For example, a transducer operating at 5 MHz produces sound with a wavelength of 0.308 mm. If each pulse consists of three cycles of sound, the pulse length is slightly less than 1 mm, and this becomes the maximum resolution along the beam axis. If the transducer frequency is increased to 15 MHz, the pulse length is less than 0.4 mm, permitting resolution of smaller details. In addition to axial resolution, resolution in the planes perpendicular to the beam axis must also be considered. Lateral resolution refers to resolution in the plane perpendicular to the beam and parallel to the transducer and is determined by the width of the ultrasound beam. Azimuth resolution, or elevation resolution, refers to the slice thickness in the plane perpendicular to the beam and to the transducer (Fig. 1-23). Ultrasound is a tomographic method of imaging that produces thin

In ultrasound, perhaps more than in any other imaging method, the quality of the information obtained is determined by the user’s ability to recognize and avoid artifacts and pitfalls.7 Many imaging artifacts are induced by errors in scanning technique or improper use of the instrument and are preventable. Artifacts may cause misdiagnosis or may obscure important findings. Understanding artifacts is essential for correct interpretation of ultrasound examinations. Many artifacts suggest the presence of structures not actually present. These include reverberation, refraction, and side lobes. Reverberation artifacts arise when the ultrasound signal reflects repeatedly between highly reflective interfaces that are usually, but not always, near the transducer (Fig. 1-24). Reverberations may also give the false impression of solid structures in areas where only fluid is present. Certain types of reverberation may be helpful because they allow the identification of a specific type of reflector, such as a surgical clip. Reverberation artifacts can usually be reduced or eliminated by changing the scanning angle or transducer placement to avoid the parallel interfaces that contribute to the artifact. Refraction causes bending of the sound beam so that targets not along the axis of the transducer are insonated. Their reflections are then detected and displayed in the image. This may cause structures to appear in the image that actually lie outside the volume the investigator assumes is being examined (see Fig. 1-7). Similarly, side lobes may produce confusing echoes that arise from sound beams that lie outside the main ultrasound beam (Fig. 1-25). These side lobe artifacts are of clinical importance because they may create the impression of structures or debris in fluid-filled structures (Fig. 1-26). Side lobes may also result in errors of measurement by reducing lateral resolution. As with most other artifacts, repositioning the transducer and its focal zone or using a different transducer will usually allow the differentiation of artifactual from true echoes.

18   PART I  ■  Physics

FIGURE 1-22.  Axial resolution. Axial resolution is the resolution along the beam axis (A) and is determined by the pulse length (B). The pulse length is the product of the wavelength (which decreases with increasing frequency) and the number of waves (usually two to three). Because the pulse length determines axial resolution, higher transducer frequencies provide higher image resolution. In B, for example, a transducer operating at 5 MHz produces sound with a wavelength of 0.31 mm. If each pulse consists of three cycles of sound, the pulse length is slightly less than 1 mm, and objects A and B, which are 0.5 mm apart, cannot be resolved as separate structures. If the transducer frequency is increased to 15 MHz, the pulse length is less than 0.3 mm, permitting A and B to be identified as separate structures.

FIGURE 1-23.  Lateral and elevation resolution. Resolution in the planes perpendicular to the beam axis is an important determinant of image quality. Lateral resolution (L) is resolution in the plane perpendicular to the beam and parallel to the transducer and is determined by the width of the ultrasound beam. Lateral resolution is controlled by focusing the beam, usually by electronic phasing to alter the beam width at a selected depth of interest. Azimuth or elevation resolution (E) is determined by the slice thickness in the plane perpendicular to the beam and the transducer. Elevation resolution is controlled by the construction of the transducer. Both lateral and elevation resolution are less than the axial resolution.

A

B

Chapter 1  ■  Physics of Ultrasound   19

FIGURE 1-24.  Reverberation artifact. Reverberation artifacts arise when the ultrasound signal reflects repeatedly between highly reflective interfaces near the transducer, resulting in delayed echo return to the transducer. This appears in the image as a series of regularly spaced echoes at increasing depth. The echo at depth 1 is produced by simple reflection from a strong interface. Echoes at levels 2, 3, and 4 are produced by multiple reflections between this interface and the surface (simulated image).

FIGURE 1-25.  Side lobes. Although most of the energy generated by a transducer is emitted in a beam along the central axis of the transducer (A), some energy is also emitted at the periphery of the primary beam (B and C). These are called side lobes and are lower in intensity than the primary beam. Side lobes may interact with strong reflectors that lie outside of the scan plane and produce artifacts that are displayed in the ultrasound image (see also Fig. 1-26).

FIGURE 1-26.  Side lobe artifact. Transverse image of the gallbladder reveals a bright internal echo (A) that suggests a band or septum within the gallbladder. This is a side lobe artifact related to the presence of a strong out-of-plane reflector (B) medial to the gallbladder. The low-level echoes in the dependent portion of the gallbladder (C) are also artifactual and are caused by the same phenomenon. Side lobe and slice thickness artifacts are of clinical importance because they may create the impression of debris in fluid-filled structures. As with most other artifacts, repositioning the transducer and its focal zone or using a different transducer will usually allow the differentiation of artifactual from true echoes.

20   PART I  ■  Physics

Artifacts may also remove real echoes from the display or obscure information, and important pathology may be missed. Shadowing results when there is a marked reduction in the intensity of ultrasound deep to a strong reflector or attenuator. Shadowing causes partial or complete loss of information due to attenuation of the sound by superficial structures. Another common cause of loss of image information is improper adjustment of system gain and TGC settings. Many low-level echoes are near the noise levels of the equipment, and considerable skill and experience are needed to adjust instrument settings to display the maximum information with the minimum noise. Poor scanning angles, inadequate penetration, and poor resolution may also result in loss of significant information. Careless selection of transducer frequency and lack of attention to the focal characteristics of the beam will cause loss of clinically important information from deep, low-amplitude reflectors and small targets. Ultrasound artifacts may alter the size, shape, and position of structures. For example, a multipath artifact is created when the path of the returning echo is not the one expected, resulting in display of the echo at an improper location in the image (Fig. 1-27).

Shadowing and Enhancement Although most artifacts degrade the ultrasound image and impede interpretation, two artifacts of clinical value are shadowing and enhancement. Again, shadowing results when an object (e.g., calculus) attenuates sound more rapidly than surrounding tissues. Enhancement occurs when an object (e.g., cyst) attenuates less than surrounding tissues. Failure of TGC applied to normal tissue to compensate properly for the attenuation of more highly attenuating (shadowing) or poorly attenuating (enhancing) structures produces the artifact (Fig. 1-28). Because attenuation increases with frequency, the effects of shadowing and enhancement are greater at higher than at lower frequencies. The conspicuity of

A

shadowing and enhancement is reduced by excessive beam width, improper focal zone placement, and use of spatial compounding.

DOPPLER SONOGRAPHY Conventional B-mode ultrasound imaging uses pulseecho transmission, detection, and display techniques. Brief pulses of ultrasound energy emitted by the transducer are reflected from acoustic interfaces within the body. Precise timing allows determination of the depth from which the echo originates. When pulsed wave ultrasound is reflected from an interface, the backscattered (reflected) signal contains amplitude, phase, and frequency information (Fig. 1-29). This information permits inference of the position, nature, and motion of the interface reflecting the pulse. B-mode ultrasound imaging uses only the amplitude information in the backscattered signal to generate the image, with differences in the strength of reflectors displayed in the image in varying shades of gray. Rapidly moving targets, such as red cells in the bloodstream, produce echoes of low amplitude that are not usually displayed, resulting in a relatively anechoic pattern within the lumens of large vessels. Although gray-scale display relies on the amplitude of the backscattered ultrasound signal, additional information is present in the returning echoes that can be used to evaluate the motion of moving targets. When highfrequency sound impinges on a stationary interface, the reflected ultrasound has essentially the same frequency or wavelength as the transmitted sound (Fig. 1-30, A). If the reflecting interface is moving with respect to the sound beam emitted from the transducer, however, there is a change in the frequency of the sound scattered by the moving object (Fig. 1-30, B and C). This change in frequency is directly proportional to the velocity of the reflecting interface relative to the transducer and is a

B

FIGURE 1-27.  Multipath artifact. A, Mirror image of the uterus is created by reflection of sound from an interface produced by gas in the rectum. B, Echoes reflected from the wall of an ovarian cyst create complex echo paths that delay return of echoes to the transducer. In both examples, the longer path of the reflected sound results in the display of echoes at a greater depth than they should normally appear. In A this results in an artifactual image of the uterus appearing in the location of the rectum. In B the effect is more subtle and more likely to cause misdiagnosis because the artifact suggests a mural nodule in what is actually a simple ovarian cyst.

Chapter 1  ■  Physics of Ultrasound   21 –0 dB

Uncorrected

Gain compensated

–10 dB

–25 dB

+10 dB

+10–25 = –15 db

–20 dB

–30 dB

+20 dB

+10–35 = –15 db

A –30 dB –0 dB

–10 + 10 dB

C

–0 dB

–20 + 20 dB

B –40 dB

+30 dB

+30–5 = –15 dB

Gain compensated

+10–3 = +7 dB

+20–13 = +7 dB

result of the Doppler effect. The relationship of the returning ultrasound frequency to the velocity of the reflector is described by the Doppler equation, as follows: ∆F = (FR − FT ) = 2 ⋅ FT ⋅ v c   5 The Doppler frequency shift is ΔF; FR is the frequency of sound reflected from the moving target; FT is the frequency of sound emitted from the transducer; v is the velocity of the target toward the transducer; and c is the velocity of sound in the medium. The Doppler frequency shift (ΔF), as just described, applies only if the target is moving directly toward or away from the transducer (Fig. 1-31, A). In most clinical settings the direction of the ultrasound beam is seldom directly toward or

FIGURE 1-28.  Shadowing and enhancement. A, Uncorrected image of a shadowing breast mass shows that the mass attenuates 15 dB more than the surrounding normal tissue. B, Application of appropriate TGC results in proper display of the normal breast tissue. However, because of the increased attenuation of the mass, a shadow results. C, Similarly, the cyst attenuates 7 dB less than the normal tissue, and TGC correction for normal tissue results in overamplification of the signals deep to the cyst, producing enhancement of these tissues.

away from the direction of flow, and the ultrasound beam usually approaches the moving target at an angle designated as the Doppler angle (Fig. 1-31, B). In this case, ΔF is reduced in proportion to the cosine of this angle, as follows: ∆F = (FR − FT ) = 2 ⋅ FT ⋅ v ⋅ cosθ c   6 where θ is the angle between the axis of flow and the incident ultrasound beam. If the Doppler angle can be measured, estimation of flow velocity is possible. Accurate estimation of target velocity requires precise measurement of both the Doppler frequency shift and the angle of insonation to the direction of target movement. As the Doppler angle (θ) approaches 90 degrees, the

22   PART I  ■  Physics

FT

FIGURE 1-29.  Backscattered information. The backscattered ultrasound signal contains amplitude, phase, and frequency information. Signals B and C differ in amplitude but have the same frequency. Amplitude differences are used to generate B-mode images. Signals A and B differ in frequency but have similar amplitudes. Such frequency differences are the basis of Doppler ultrasound.

FR

v ∆F = (FR − FT) = 2 • FT • v c

A A

Stationary target: (FR − FT) = 0

B

Target motion toward transducer: (FR − FT) > 0

FT

FR

θ

C

Target motion away from transducer: (FR − FT) < 0

FIGURE 1-30.  Doppler effect. A, Stationary target. If the reflecting interface is stationary, the backscattered ultrasound has the same frequency or wavelength as the transmitted sound, and there is no difference in the transmitted (FT) and reflected (FR) frequencies. B and C, Moving targets. If the reflecting interface is moving with respect to the sound beam emitted from the transducer, there is a change in the frequency of the sound scattered by the moving object. When the interface moves toward the transducer (B), the difference in reflected and transmitted frequencies is greater than zero. When the target is moving away from the transducer (C), this difference is less than zero. The Doppler equation is used to relate this change in frequency to the velocity of the moving object.

cosine of θ approaches 0. At an angle of 90 degrees, there is no relative movement of the target toward or away from the transducer, and no Doppler frequency shift is detected (Fig. 1-32). Because the cosine of the Doppler angle changes rapidly for angles more than 60

v

B

∆F = (FR − FT) = 2 • FT • v • cos θ c

FIGURE 1-31.  Doppler equations. The Doppler equation describes the relationship of the Doppler frequency shift to target velocity. A, In its simplest form, it is assumed that the direction of the ultrasound beam is parallel to the direction of movement of the target. This situation is unusual in clinical practice. More often, the ultrasound impinges on the vessel at angle θ. B, In this case the Doppler frequency shift detected is reduced in proportion to the cosine of θ.

degrees, accurate angle correction requires that Doppler measurements be made at angles of less than 60 degrees. Above 60 degrees, relatively small changes in the Doppler angle are associated with large changes in cosθ, and therefore a small error in estimation of the Doppler angle

Chapter 1  ■  Physics of Ultrasound   23

θ = 60° cos θ = 0.5 ∆F = 0.5

θ = 90° cos θ = 0.0 ∆F = 0.0 B

A θ = 0° cos θ = 1.0 ∆F = 1.0

FIGURE 1-32.  Effect of Doppler angle on frequency shift. At an angle of 60 degrees, the detected frequency shift detected by the transducer is only 50% of the shift detected at an angle of 0 degrees. At 90 degrees, there is no relative movement of the target toward or away from the transducer, and no frequency shift is detected. The detected Doppler frequency shift is reduced in proportion to the cosine of the Doppler angle. Because the cosine of the angle changes rapidly at angles above 60 degrees, the use of Doppler angles of less than 60 degrees is recommended in making velocity estimates.

may result in a large error in the estimation of velocity. These considerations are important in using both duplex and color Doppler instruments. Optimal imaging of the vessel wall is obtained when the axis of the transducer is perpendicular to the wall, whereas maximal Doppler frequency differences are obtained when the transducer axis and the direction of flow are at a relatively small angle. In peripheral vascular applications, it is highly desirable that measured Doppler frequencies be corrected for the Doppler angle to provide velocity measurement. This allows comparison of data from systems using different Doppler frequencies and eliminates error in interpretation of frequency data obtained at different Doppler angles. For abdominal applications, anglecorrected velocity measurements are encouraged, although qualitative assessments of flow are often made using only the Doppler frequency shift data. The interrelation of transducer frequency (FT) and the Doppler angle (θ) to the Doppler frequency shift (ΔF) and target velocity described by the Doppler equation are important in proper clinical use of Doppler equipment.

Doppler Signal Processing and Display Several options exist for the processing of ΔF, the Doppler frequency shift, to provide useful information regarding the direction and velocity of blood. Doppler frequency shifts encountered clinically are in the audible

FIGURE 1-33.  Doppler display. A, Doppler frequency spectrum waveform shows changes in flow velocity and direction by vertical deflections of the waveform above and below the baseline. The width of the spectral waveform (spectral broadening) is determined by the range of frequencies present at any instant in time (arrow). A brightness (gray) scale is used to indicate the amplitude of each frequency component. B, Color Doppler imaging. Amplitude data from stationary targets provide the basis for the B-mode image. Signal phase provides information about the presence and direction of motion, and changes in frequency relate to the velocity of the target. Backscattered signals from red blood cells are displayed in color as a function of their motion toward or away from the transducer, and the degree of the saturation of the color is used to indicate the frequency shift from moving red cells.

range. This audible signal may be analyzed by ear and, with training, the operator can identify many flow characteristics. More often, the Doppler shift data are displayed in graphic form as a time-varying plot of the frequency spectrum of the returning signal. A fast Fourier transformation is used to perform the frequency analysis. The resulting Doppler frequency spectrum displays the following (Fig. 1-33, A): • Variation with time of the Doppler frequencies present in the volume sampled. • The envelope of the spectrum, representing the maximum frequencies present at any given point in time. • The width of the spectrum at any point, indicating the range of frequencies present. The amplitude of the Doppler signal is related to the number of targets moving at a given velocity. In many instruments the amplitude of each frequency component is displayed in gray scale as part of the spectrum. The presence of a large number of different frequencies at a given point in the cardiac cycle results in spectral broadening. In color Doppler imaging systems, a representation of the Doppler frequency shift is displayed as a feature of the image itself (Fig. 1-33, B). In addition to the detection of Doppler frequency shift data from each pixel in the image, these systems may also provide range-gated pulsed wave Doppler with spectral analysis for display of Doppler data.

24   PART I  ■  Physics

Doppler Instrumentation In contrast to A-mode, M-mode, and B-mode gray-scale ultrasonography, which display the information from tissue interfaces, Doppler ultrasound instruments are optimized to display flow information. The simplest Doppler devices use continuous wave rather than pulsed wave ultrasound, using two transducers that transmit and receive ultrasound continuously (continuous wave or CW Doppler). The transmit and receive beams overlap in a sensitive volume at some distance from the transducer face (Fig. 1-34, A). Although direction of flow can be determined with CW Doppler, these devices do not allow discrimination of motion coming from various depths, and the source of the signal being detected is difficult, if not impossible, to ascertain with certainty. Inexpensive and portable, CW Doppler instruments are used primarily at the bedside or intraoperatively to confirm the presence of flow in superficial vessels. Because of the limitations of CW systems, most applications use range-gated, pulsed wave Doppler. Rather than a continuous wave of ultrasound emission, pulsed wave Doppler devices emit brief pulses of ultrasound energy (Fig. 1-34, B). Using pulses of sound permits use of the time interval between the transmission of a pulse and the return of the echo as a means of determining the depth from which the Doppler shift arises. The principles are similar to the echo-ranging principles used for imaging (see Fig. 1-4). In a pulsed wave Doppler system the sensitive volume from which flow data are sampled can be controlled in terms of shape, depth, and position. When pulsed wave Doppler is combined with a 2-D,

A

real-time, B-mode imager in the form of a duplex scanner, the position of the Doppler sample can be precisely controlled and monitored. The most common form of Doppler ultrasound to be used for radiology applications is color Doppler imaging8 (Fig. 1-35, A). In color Doppler imaging systems, frequency shift information determined from Doppler measurements is displayed as a feature of the image itself. Stationary or slowly moving targets provide the basis for the B-mode image. Signal phase provides information about the presence and direction of motion, and changes in echo signal frequency relate to the velocity of the target. Backscattered signals from red blood cells are displayed in color as a function of their motion toward or away from the transducer, and the degree of the saturation of the color is used to indicate the relative frequency shift produced by the moving red cells. Color Doppler flow imaging (CDFI) expands conventional duplex sonography by providing additional capabilities. The use of color saturation to display variations in Doppler shift frequency allows an estimation of relative velocity from the image alone, provided that variations in the Doppler angle are noted. The display of flow throughout the image field allows the position and orientation of the vessel of interest to be observed at all times. The display of spatial information with respect to velocity is ideal for display of small, localized areas of turbulence within a vessel, which provide clues to stenosis or irregularity of the vessel wall caused by atheroma, trauma, or other disease. Flow within the vessel is observed at all points, and stenotic jets and focal areas of turbulence are displayed that might be overlooked with duplex instrumentation. The contrast of

B

FIGURE 1-34.  Continuous wave and pulsed wave Doppler. A, Continuous wave (CW) Doppler uses separate transmit and receive crystals that continuously transmit and receive ultrasound. Although able to detect the presence and direction of flow, CW devices are unable to distinguish signals arising from vessels at different depths (green-shaded area). B, Using the principle of ultrasound ranging (see Fig. 1-4), pulsed wave Doppler permits the sampling of flow data from selected depths by processing only the signals that return to the transducer after precisely timed intervals. The operator is able to control the position of the sample volume and, in duplex systems, to view the location from which the Doppler data are obtained.

Chapter 1  ■  Physics of Ultrasound   25

A

B

FIGURE 1-35.  Color flow and power mode Doppler. A, Color flow Doppler imaging uses a color map to display information based on the detection of frequency shifts from moving targets. Noise in this form of display appears across the entire frequency spectrum and limits sensitivity. B, Power mode Doppler uses a color map to show the distribution of the power or amplitude of the Doppler signal. Flow direction and velocity information are not provided in power mode Doppler display, but noise is reduced, allowing higher gain settings and improved sensitivity for flow detection.

LIMITATIONS OF COLOR DOPPLER FLOW IMAGING Angle dependence Aliasing Inability to display entire Doppler spectrum in the image Artifacts caused by noise

flow within the vessel lumen (1) permits visualization of small vessels that are invisible when using conventional imagers and (2) enhances the visibility of wall irregularity. CDFI aids in determination of the direction of flow and measurement of the Doppler angle.

Power Mode Doppler An alternative to the display of frequency information with color Doppler imaging is to use a color map that displays the integrated power of the Doppler signal instead of its mean frequency shift9 (Fig. 1-35, B). Because frequency shift data are not displayed, there is no aliasing. The image does not provide information related to flow direction or velocity, and power mode Doppler imaging is much less angle dependent than frequency-based color Doppler display. In contrast to color Doppler, where noise may appear in the image as any color, power mode Doppler permits noise to be assigned to a homogeneous background color that does not greatly interfere with the image. This results in a significant increase in the usable dynamic range of the scanner, permitting higher effective gain settings and increased sensitivity for flow detection (Fig. 1-36).

ADVANTAGES OF POWER MODE DOPPLER No aliasing Much less angle dependence Noise: a homogeneous background color Increased sensitivity for flow detection

Interpretation of the Doppler Spectrum Doppler data components that must be evaluated both in spectral display and in color Doppler imaging include the Doppler shift frequency and amplitude, the Doppler angle, the spatial distribution of frequencies across the vessel, and the temporal variation of the signal. Because the Doppler signal itself has no anatomic significance, the examiner must interpret the Doppler signal and then determine its relevance in the context of the image. The detection of a Doppler frequency shift indicates movement of the target, which in most applications is related to the presence of flow. The sign of the frequency shift (positive or negative) indicates the direction of flow relative to the transducer. Vessel stenosis is typically associated with large Doppler frequency shifts in both systole and diastole at the site of greatest narrowing, with turbulent flow in poststenotic regions. In peripheral vessels, analysis of the Doppler changes allows accurate prediction of the degree of vessel narrowing. Information related to the resistance to flow in the distal vascular tree can be obtained by analysis of changes of blood velocity with time, as shown in the Doppler spectral display.

26   PART I  ■  Physics

FIGURE 1-36.  Frequency and power mode color mapping. A, Conventional color Doppler uses the color map to show differences in flow direction and Doppler frequency shift. Because noise appears over the entire frequency spectrum, gain levels are limited to those that do not introduce excessive noise. B, Power mode Doppler color map, in contrast, indicates the amplitude of the Doppler signal. Because most noise is of low amplitude, it is possible to map this to colors near the background. This permits the use of high gain settings that offer significant improvements over conventional color Doppler in flow detection.

A

B

Figure 1-37 provides a graphic example of the changes in the Doppler spectral waveform resulting from physiologic changes in the resistance of the vascular bed supplied by a normal brachial artery. A blood pressure cuff has been inflated to above systolic pressure to occlude the distal branches supplied by the brachial artery. This occlusion causes reduced systolic amplitude and cessation of diastolic flow, resulting in a waveform different than that found in the normal resting state. During the period of ischemia induced by pressure cuff occlusion of the forearm vessels, vasodilation has occurred. The Doppler waveform now reflects a low-resistance peripheral vascular bed with increased systolic amplitude and rapid flow throughout diastole, typical for vasodilation. Doppler indices include the systolic/diastolic ratio, resistive index, and pulsatility index (Fig. 1-38). These compare blood flow in systole and diastole, show resistance to flow in the peripheral vascular bed, and help evaluate the perfusion of tumors, renal transplants, the placenta, and other organs. With Doppler ultrasound, it is therefore possible to identify vessels, determine the direction of blood flow, evaluate narrowing or occlusion, and characterize blood flow to organs and tumors. Analysis of the Doppler shift frequency with time can be used to infer both proximal stenosis and changes in distal vascular impedance. Most work using pulsed wave Doppler imaging has emphasized the detection of stenosis, thrombosis, and flow disturbances in major periph-

eral arteries and veins. In these applications, measurement of peak systolic and end diastolic frequency or velocity, analysis of the Doppler spectrum, and calculation of certain frequency or velocity ratios have been the basis of analysis. Changes in the spectral waveform measured by indices comparing flow in systole and diastole indicate the resistance of the vascular bed supplied by the vessel and the changes resulting from a variety of pathologies. Changes in Doppler indices from normal may help in the early identification of rejection of transplanted organs, parenchymal dysfunction, and malignancy. Although useful, these measurements are influenced not only by the resistance to flow in peripheral vessels, but also by heart rate, blood pressure, vessel wall length and elasticity, extrinsic organ compression, and other factors. Therefore, interpretation must always take into account these variables.

Interpretation of Color Doppler Although the graphic presentation of color Doppler imaging suggests that interpretation is made easier, the complexity of the color Doppler image actually makes this a more demanding image to evaluate than the simple Doppler spectrum. Nevertheless, color Doppler imaging has important advantages over pulsed wave duplex Doppler imaging, in which flow data are obtained only from a small portion of the area being imaged. To be

Chapter 1  ■  Physics of Ultrasound   27

FIGURE 1-37.  Impedance. A, High-resistance waveform in brachial artery produced by inflating forearm blood pressure cuff to a pressure above the systolic blood pressure. As a result of high peripheral resistance, there is low systolic amplitude and reversed diastolic flow. B, Low-resistance waveform in peripheral vascular bed caused by vasodilation stimulated by the prior ischemia. Immediately after release of 3 minutes of occluding pressure, the Doppler waveform showed increased amplitude and rapid antegrade flow throughout diastole.

FIGURE 1-38.  Doppler indices. Doppler imaging can provide information about blood flow in both large and small vessels. Small vessel impedance is reflected in the Doppler spectral waveform of afferent vessels. Doppler flow indices used to characterize peripheral resistance are based on the peak systolic frequency or velocity (A), the minimum or end diastolic frequency or velocity (B), and the mean frequency or velocity (M). The most frequently used indices are the systolic/diastolic ratio (A/B); resistive index [(A-B)/A]; and pulsatility index [(A-B)/ M]. In calculation of the pulsatility index, the minimum diastolic velocity or frequency is used; calculation of the systolic/diastolic ratio and resistive index use the end diastolic value.

confident that a conventional Doppler study has achieved reasonable sensitivity and specificity in detection of flow disturbances, a methodical search and sampling of multiple sites within the field of interest must be performed. In contrast, CDFI devices permit simultaneous sampling of multiple sites and are less susceptible to this error.

Although color Doppler can indicate the presence of blood flow, misinterpretation of color Doppler images may result in significant errors. Each color pixel displays a representation of the Doppler frequency shift detected at that point. The frequency shift displayed is not the peak frequency present at sampling but rather a weighted mean frequency that attempts to account for the range of frequencies and their relative amplitudes at sampling. Manufacturers use different methods to derive the weighted mean frequency displayed in their systems. In addition, the pulse repetition frequency (PRF) and the color map selected to display the detected range of frequencies affect the color displayed. The color assigned to each Doppler pixel is determined by the Doppler frequency shift (which in turn is determined by target velocity and Doppler angle), the PRF, and the color map selected for display; therefore the interpretation of a color Doppler image must consider each of these variables. Although most manufacturers provide on-screen indications suggesting a relationship between the color displayed and flow velocity, this is misleading because color Doppler does not show velocity and only indicates the weighted mean frequency shift measured in the vessel; without correction for the effect of the Doppler angle, velocity cannot be estimated (Fig. 1-39). Since the frequency shift at a given point is a function of velocity and the Doppler angle, depending on the frequency shift present in a given pixel and the PRF, any velocity may be represented by any color, and under certain circumstances, low-velocity flow may not be shown at all. As with spectral Doppler, aliasing is determined by PRF.

28   PART I  ■  Physics

FIGURE 1-39.  Color Doppler. Each color pixel in a color Doppler image represents the Doppler frequency shift at that point, and it cannot be used to estimate velocity. Even though the points A and B have similar color values and therefore similar Doppler frequencies, the velocity at A is much higher than at B because of the large Doppler angle at A compared to B. The velocity represented by a given Doppler frequency increases in proportion to the Doppler angle.

A (PRF = 700 Hz)

A

B

B (PRF = 4500 Hz)

FIGURE 1-40.  Pulse repetition frequency (PRF). Depending on the color map selected, velocity of the target, Doppler angle, and PRF, a given velocity may appear as any color with color Doppler. A and B are sonograms of identical vessels. A, PRF is 700 Hz, which results in aliasing of the higher Doppler frequency shifts in the carotid artery, but permits the identification of relatively slow flow in the jugular vein. B, PRF is 4500 Hz, eliminating aliasing in the artery but also suppressing the display of the low Doppler frequencies in the internal jugular vein.

With color Doppler, aliasing causes frequencies greater than twice the PRF to “wrap around” and to be displayed in the opposite colors of the color map. Inexperienced users tend to associate color Doppler aliasing with elevated velocity, but even low velocities may show marked aliasing if PRF is sufficiently low. As PRF is increased, aliasing of high Doppler frequency shifts is reduced; however, low frequency shifts may be eliminated from the display, resulting in diagnostic error (Fig. 1-40).

Other Technical Considerations Although many problems and artifacts associated with B-mode imaging (e.g., shadowing) are encountered with Doppler sonography, the detection and display of frequency information related to moving targets present additional technical considerations. It is important to

understand the source of these artifacts and their influence on the interpretation of the flow measurements obtained in clinical practice.

Doppler Frequency A primary objective of the Doppler examination is the accurate measurement of characteristics of flow within a vascular structure. The moving red blood cells that serve as the primary source of the Doppler signal act as point scatterers of ultrasound rather than specular reflectors. This interaction results in the intensity of the scattered sound varying in proportion to the fourth power of the frequency, which is important in selecting the Doppler frequency for a given examination. As the transducer frequency increases, Doppler sensitivity improves, but attenuation by tissue also increases, resulting in dimin-

Chapter 1  ■  Physics of Ultrasound   29

MAJOR SOURCES OF DOPPLER IMAGING ARTIFACTS DOPPLER FREQUENCY

Higher frequencies lead to more tissue attenuation. Wall filters remove signals from low-velocity blood flow.

INCREASE IN SPECTRAL BROADENING

Excessive system gain or changes in dynamic range of the gray-scale display. Excessively large sample volume. Sample volume too near the vessel wall.

INCREASE IN ALIASING

Decrease in pulse repetition frequency PRF. Decrease in Doppler angle. Higher Doppler frequency transducer.

DOPPLER ANGLE

Relatively inaccurate above 60 degrees.

Wall Filters Doppler instruments detect motion not only from blood flow but also from adjacent structures. To eliminate these low-frequency signals from the display, most instruments use high pass filters, or “wall” filters, which remove signals that fall below a given frequency limit. Although effective in eliminating low-frequency noise, these filters may also remove signals from low-velocity blood flow (Fig. 1-41). In certain clinical situations the measurement of these slower flow velocities is of clinical importance, and the improper selection of the wall filter may result in serious errors of interpretation. For example, low-velocity venous flow may not be detected if an improper filter is used, and low-velocity diastolic flow in certain arteries may also be eliminated from the display, resulting in errors in the calculation of Doppler indices, such as the systolic/diastolic ratio or resistive index. In general, the filter should be kept at the lowest practical level, usually 50 to 100 Hz.

SAMPLE VOLUME SIZE

Large sample volumes increase vessel wall noise.

ished penetration. Careful balancing of the requirements for sensitivity and penetration is an important responsibility of the operator during a Doppler examination. Because many abdominal vessels lie several centimeters beneath the surface, Doppler frequencies in the range of 3 to 3.5 MHz are usually required to permit adequate penetration.

A

B

Spectral Broadening Spectral broadening refers to the presence of a large range of flow velocities at a given point in the pulse cycle and, by indicating turbulence, is an important criterion of high-grade vessel narrowing. Excessive system gain or changes in the dynamic range of the gray-scale display of the Doppler spectrum may suggest spectral broadening; opposite settings may mask broadening of the Doppler spectrum, causing diagnostic inaccuracy. Spectral broadening may also be produced by the selection

FIGURE 1-41.  Wall fil­ ters. Wall filters are used to eliminate low-frequency noise from the Doppler display. Here the effect on the display of low-velocity flow is shown with wall filter settings of A, 100 Hz, and B, 400 Hz. High wall filter settings remove signal from low-velocity blood flow and may result in interpretation errors. In general, wall filters should be kept at the lowest practical level, usually in the range of 50 to 100 Hz.

30   PART I  ■  Physics

mitting the next pulse. This limits the rate with which pulses can be generated, a lower PRF being required for greater depth. The PRF also determines the maximum depth from which unambiguous data can be obtained. If PRF is less than twice the maximum frequency shift produced by movement of the target (Nyquist limit), aliasing results (Fig. 1-43, A and B). When PRF is less than twice the frequency shift being detected, lower frequency shifts than are actually present are displayed. Because of the need for lower PRFs to reach deep vessels, signals from deep abdominal arteries are prone to aliasing if high velocities are present. In practice, aliasing is usually readily recognized (Fig. 1-43, C and D). Aliasing can be reduced by increasing the PRF, by increasing the Doppler angle—thereby decreasing the frequency shift— or by using a lower-frequency Doppler transducer.

Doppler Angle When making Doppler measurement of velocity, it is necessary to correct for the Doppler angle. The accuracy of a velocity estimate obtained with Doppler is only as great as the accuracy of the measurement of the Doppler angle. This is particularly true as the Doppler angle exceeds 60 degrees. In general, the Doppler angle is best kept at 60 degrees or less because small changes in the Doppler angle above 60 degrees result in significant changes in the calculated velocity. Therefore, measurement inaccuracies result in much greater errors in velocity estimates than do similar errors at lower Doppler angles. Angle correction is not required for the measurement of Doppler indices such as the resistive index, because these measurements are based only on the relationship of the systolic and diastolic amplitudes.

Sample Volume Size FIGURE 1-42.  Spectral broadening. The range of velocities detected at a given time in the pulse cycle is reflected in the Doppler spectrum as spectral broadening. A, Normal spectrum. Spectral broadening may arise from turbulent flow in association with vessel stenosis. B and C, Artifactual spectral broadening may be produced by improper positioning of the sample volume near the vessel wall, use of an excessively large sample volume (B), or excessive system gain (C).

With pulsed wave Doppler systems, the length of the Doppler sample volume can be controlled by the operator, and the width is determined by the beam profile. Analysis of Doppler signals requires that the sample volume be adjusted to exclude as much of the unwanted clutter as possible from near the vessel walls.

Doppler Gain of an excessively large sample volume or by the placement of the sample volume too near the vessel wall, where slower velocities are present (Fig. 1-42).

Aliasing Aliasing is an artifact arising from ambiguity in the measurement of high Doppler frequency shifts. To ensure that samples originate from only a selected depth when using a pulsed wave Doppler system, it is necessary to wait for the echo from the area of interest before trans-

As with imaging, proper gain settings are essential to accurate and reproducible Doppler measurements. Excessive Doppler gain results in noise appearing at all frequencies and may result in overestimation of velocity. Conversely, insufficient gain may result in underestimation of peak velocity (Fig. 1-44). A consistent approach to setting Doppler gain should be used. After placing the sample volume in the vessel, the Doppler gain should be increased to a level where noise is visible in the image, then gradually reduced to the point at which the noise first disappears completely.

Chapter 1  ■  Physics of Ultrasound   31

A

C

B

D

FIGURE 1-43.  Aliasing. Pulse repetition frequency (PRF) determines the sampling rate of a given Doppler frequency. A, If PRF (arrows) is sufficient, the sampled waveform (orange curve) will accurately estimate the frequency being sampled (yellow curve). B, If PRF is less than half the frequency being measured, undersampling will result in a lower frequency shift being displayed (orange curve). C, In a clinical setting, aliasing appears in the spectral display as a “wraparound” of the higher frequencies to display below the baseline. D, In color Doppler display, aliasing results in a wraparound of the frequency color map from one flow direction to the opposite direction, passing through a transition of unsaturated color. The velocity throughout the vessel is constant, but aliasing appears only in portions of the vessel because of the effect of the Doppler angle on the Doppler frequency shift. As the angle increases, the Doppler frequency shift decreases, and aliasing is no longer seen.

OPERATING MODES: CLINICAL IMPLICATIONS Ultrasound devices may operate in several modes, including real-time, color Doppler, spectral Doppler, and M-mode imaging. Imaging is produced in a scanned mode of operation. In scanned modes, pulses of ultrasound from the transducer are directed down lines of sight that are moved or steered in sequence to generate the image. This means that the number of ultrasound pulses arriving at a given point in the patient over a given interval is relatively small, and relatively little energy is deposited at any given location. In contrast, spectral Doppler imaging is an unscanned mode of operation in which multiple ultrasound pulses are sent in repetition

along a line to collect the Doppler data. In this mode the beam is stationary, resulting in considerably greater potential for heating than in imaging modes. For imaging, PRFs are usually a few thousand hertz with very short pulses. Longer pulse durations are used with Doppler than with other imaging modes. In addition, to avoid aliasing and other artifacts with Doppler imaging, it is often necessary to use higher PRFs than with other imaging applications. Longer pulse duration and higher PRF result in higher duty factors for Doppler modes of operation and increase the amount of energy introduced in scanning. Color Doppler, although a scanned mode, produces exposure conditions between those of real-time and Doppler imaging because color Doppler devices tend to send more pulses down each scan line and may use longer pulse durations than

32   PART I  ■  Physics

A

Excess gain PSV = 75 cm/sec

B

Proper gain PSV = 60 cm/sec

C

Insufficent gain PSV = 50 cm/sec

FIGURE 1-44.  Doppler gain. Accurate estimation of velocity requires proper Doppler gain adjustment. Excessive gain will cause an overestimation of peak velocity (A), and insufficient gain will result in underestimation of velocity (C). To adjust gain properly, the sample volume and Doppler angle are first set at the sample site. The gain is turned up until noise appears in the background (A), then is gradually reduced just to the point where the background noise disappears from the image (B).

imaging devices. Clearly, every user needs to be aware that switching from an imaging to a Doppler mode changes the exposure conditions and the potential for biologic effects (bioeffects). With current devices operating in imaging modes, concerns about bioeffects are minimal because intensities sufficient to produce measurable heating are seldom used. With Doppler ultrasound, the potential for thermal effects is greater. Preliminary measurements on commercially available instruments suggest that at least some of these instruments are capable of producing temperature rises of greater than 1° C at soft tissue/bone interfaces, if the focal zone of the transducer is held stationary. Care is therefore warranted when Doppler measurements are obtained at or near soft tissue/bone interfaces, as in the second and third trimester of pregnancy. These applications require thoughtful application of the principle of ALARA (as low as reasonably achievable). Under ALARA the user should use the lowest possible acoustic exposure to obtain the necessary diagnostic information.

Bioeffects and User Concerns Although users of ultrasound need to be aware of bioeffects concerns, another key factor to consider in the safe use of ultrasound is the user. The knowledge and skill of the user are major determinants of the risk-to-benefit implications of the use of ultrasound in a specific clinical situation. For example, an unrealistic emphasis on risks may discourage an appropriate use of ultrasound, resulting in harm to the patient by preventing the acquisition of useful information or by subjecting the patient to

another, more hazardous examination. The skill and experience of the individual performing and interpreting the examination are likely to have a major impact on the overall benefit of the examination. In view of the rapid growth of ultrasound and its proliferation into the hands of minimally trained clinicians, many more patients are likely to be harmed by misdiagnosis resulting from improper indications, poor examination technique, and errors in interpretation than from bioeffects. Failure to diagnose a significant anomaly or misdiagnosis (e.g., of ectopic pregnancy) are real dangers, and poorly trained users may be the greatest current hazard of diagnostic ultrasound. Understanding bioeffects is essential for the prudent use of diagnostic ultrasound and is important in ensuring that the excellent risk-to-benefit performance of diagnostic ultrasound is preserved. All users of ultrasound should be prudent, understanding as fully as possible the potential risks and obvious benefits of ultrasound examinations, as well as those of alternate diagnostic methods. With this information, operators can monitor exposure conditions and implement the principle of ALARA to keep patient and fetal exposure as low as possible while fulfilling diagnostic objectives.

THERAPEUTIC APPLICATIONS: HIGH-INTENSITY FOCUSED ULTRASOUND Although the primary medical application of ultrasound has been for diagnosis, therapeutic applications are developing rapidly, particularly the use of high-intensity focused ultrasound (HIFU). HIFU is based on three important capabilities of ultrasound: (1) focusing the ultrasound beam to produce highly localized energy deposition, (2) controlling the location and size of the focal zone, and (3) using intensities sufficient to destroy tissue at the focal zone. This has led to an interest in HIFU as a means of destroying noninvasive tumor and controlling bleeding and cardiac conduction anomalies. High-intensity focused ultrasound exploits thermal (heating of tissues) and mechanical (cavitation) bioeffect mechanisms. As ultrasound passes through tissue, attenuation occurs through scattering and absorption. Scattering of ultrasound results in the return of some of the transmitted energy to the transducer, where it is detected and used to produce an image, or Doppler display. The remaining energy is transmitted to the molecules in the acoustic field and produces heating. At the spatial peak temporal average (SPTA), intensities of 50 to 500 mW/ cm2 used for imaging and Doppler, heating is minimal, and no observable bioeffects related to tissue heating in humans have yet been documented with clinical devices. With higher intensities, however, tissue heating sufficient to destroy tissue may be achieved. Using HIFU at 1 to 3 mHz, focal peak intensities of 5000 to 20,000 W/

Chapter 1  ■  Physics of Ultrasound   33

by bowel gas, aerated lung, or bone may result in tissue heating along the reflected path of the sound, producing unintended tissue damage. Major challenges with HIFU include image guidance and accurate monitoring of therapy as it is being delivered. Magnetic resonance imaging (MRI) provides a means of monitoring temperature elevation during treatment, which is not possible with ultrasound. Guidance of therapy may be done with ultrasound or MRI, with ultrasound guidance having the advantage of verification of the acoustic window and sound path for the delivery of HIFU.

FIGURE 1-45.  High-intensity focused ultrasound (HIFU). Local tissue destruction by heating may be achieved using HIFU delivered with focal peak intensities of several thousand W/cm2. Tissue destruction can be confined to a small area a few millimeters in size without injury to adjacent tissues. HIFU is a promising tool for minimally invasive treatment of bleeding sites, uterine fibroids, and tumors in the prostate, liver, and breast.

cm2 may be achieved. This energy can be delivered to a small point several millimeters in size, producing rapid temperature elevation and resulting in tissue coagulation, with little damage to adjacent tissues (Fig. 1-45). The destruction of tissue is a function of the temperature reached and the duration of the temperature elevation. In general, elevation of tissue to a temperature of 60° C for 1 second is sufficient to produce coagulation necrosis. These conditions are readily achieved with HIFU. Because of its ability to produce highly localized tissue destruction, HIFU has been investigated as a tool for noninvasive or minimally invasive treatment of bleeding sites, uterine fibroids, and tumors in the prostate, liver, and breast.10,11 As with diagnostic ultrasound, HIFU is limited by the presence of gas or bone interposed between the transducer and the target tissue. The reflection of high-energy ultrasound from strong interfaces produced

References Basic Acoustics 1. Chivers RC, Parry RJ. Ultrasonic velocity and attenuation in mammalian tissues. J Acoust Soc Am 1978;63:940-953. 2. Goss SA, Johnston RL, Dunn F. Comprehensive compilation of empirical ultrasonic properties of mammalian tissues. J Acoust Soc Am 1978;64:423-457. 3. Merritt CR, Kremkau FW, Hobbins JC. Diagnostic ultrasound: bioeffects and safety. Ultrasound Obstet Gynecol 1992;2:366-374. 4. Medical diagnostic ultrasound instrumentation and clinical interpretation. Report of the Ultrasonography Task Force, Council on Scientific Affairs. JAMA 1991;265:1155-1159. Instrumentation 5. Krishan S, Li PC, O’Donnell M. Adaptive compensation of phase and magnitude aberrations. IEEE Trans Ultrasonics Fer Freq Control 1996;43:44. 6. Merritt CR. Technology update. Radiol Clin North Am 2001;39: 385-397. 7. Merritt CR. Doppler US: the basics. Radiographics 1991;11:109119. 8. Merritt CR. Doppler color flow imaging. J Clin Ultrasound 1987;15: 591-597. 9. Rubin JM, Bude RO, Carson PL, et al. Power Doppler US: a potentially useful alternative to mean frequency-based color Doppler US. Radiology 1994;190:853-856. Therapeutic Applications: High-Intensity Focused Ultrasound 10. Dubinsky TJ, Cuevas C, Dighe MK, et al. High-intensity focused ultrasound: current potential and oncologic applications. AJR Am J Roentgenol 2008;190:191-199. 11. Kennedy JE, Ter Haar GR, Cranston D. High-intensity focused ultrasound: surgery of the future? Br J Radiol 2003;76:590-599.

CHAPTER 2 

Biologic Effects and Safety J. Brian Fowlkes and Christy K. Holland

Chapter Outline REGULATION OF ULTRASOUND OUTPUT PHYSICAL EFFECTS OF SOUND THERMAL EFFECTS Ultrasound Produces Heat Factors Controlling Tissue Heating Spatial Focusing Temporal Considerations Tissue Type

Bone Heating Soft Tissue Heating Hyperthermia and Ultrasound Safety Thermal Index Homogeneous Tissue Model (Soft Tissue)

Tissue Model with Bone at the Focus (Fetal Applications) Tissue Model with Bone at the Surface (Transcranial Applications) Estimate of Thermal Effects

Summary Statement on Thermal Effects EFFECTS OF ACOUSTIC CAVITATION Potential Sources for Bioeffects Sonochemistry Evidence of Cavitation from Lithotripters Bioeffects in Lung and Intestine Ultrasound Contrast Agents

U

ltrasound has provided a wealth of knowledge in diagnostic medicine and has greatly impacted medical practice, particularly obstetrics. Millions of sonographic examinations are performed each year, and ultrasound remains one of the fastest-growing imaging modalities because of its low cost, real-time interactions, portability, and apparent lack of biologic effects (bioeffects). No casual relationship has been established between clinical applications of diagnostic ultrasound and bioeffects on the patient or operator.

REGULATION OF ULTRASOUND OUTPUT At present, the U.S. Food and Drug Administration (FDA) regulates the maximum output of ultrasound devices to an established level. The marketing approval process requires devices to be equivalent in efficacy and output to those produced before 1976. This historic regulation of sonography has provided a safety margin for ultrasound while allowing clinically useful performance. The mechanism has restricted ultrasound exposure to levels that apparently produce few, if any, obvious bioeffects based on the epidemiologic evidence, although animal studies have shown some evidence for biologic effects. 34

Mechanical Index Summary Statement on Gas Body Bioeffects OUTPUT DISPLAY STANDARD GENERAL AIUM SAFETY STATEMENTS EPIDEMIOLOGY CONTROLLING ULTRASOUND OUTPUT ULTRASOUND ENTERTAINMENT VIDEOS

In an effort to increase the efficacy of diagnostic ultrasound, the maximum acoustic output for some applications has increased through an additional FDA market approval process termed “510K Track 3.” The vast majority of ultrasound systems currently in use were approved through this process. The Track 3 process provides the potential for better imaging performance and, as discussed later, requires that additional information be reported to the operator regarding the relative potential for bioeffects. Therefore, informed decision making is important concerning the possible adverse effects of ultrasound in relation to the desired diagnostic information. Current FDA regulations that limit the maximum output are still in place, but in the future, systems might allow sonographers and physicians the discretion to increase acoustic output beyond a level that might induce a biologic response. Although the choices made during sonographic examinations may not be equivalent to the risk-versus-benefit decisions associated with imaging modalities using ionizing radiation, the operator will be increasingly responsible for determining the diagnostically required amount of ultrasound exposure. Thus the operator should know the potential bioeffects associated with ultrasound exposure. Patients also need to be reassured about the safety of a diagnostic ultrasound scan. The scientific community has identified some potential bioeffects from sono­

Chapter 2  ■  Biologic Effects and Safety   35

graphy, and although no causal relation has been established, it does not mean that no effects exist. Therefore it is important to understand the interaction of ultrasound with biologic systems.

path of the ultrasound by absorption. Absorption loss occurs substantially through the conversion of the ultrasound energy into heat. This heating provides a mechanism for ultrasound-induced bioeffects.

PHYSICAL EFFECTS OF SOUND

Factors Controlling Tissue Heating

The physical effects of sound can be divided into two principal groups: thermal and nonthermal. Most medical professionals recognize the thermal effects of elevated temperature on tissue, and the effects caused by ultrasound are similar to those of any localized heat source. With ultrasound the heating mainly results from the absorption of the sound field as it propagates through tissue. However, “nonthermal” sources can generate heat as well. Many nonthermal mechanisms for bioeffects exist. Acoustic fields can apply radiation forces (not ionizing radiation) on the structures within the body both at the macroscopic and the microscopic level, resulting in exerted pressure and torque. The temporal average pressure in an acoustic field is different than the hydrostatic pressure of the fluid, and any object in the field is subject to this change in pressure. The effect is typically considered smaller than other effects because it relies on less significant factors in the formulation of the acoustic field. Acoustic fields can also cause motion of fluids. Such acoustically induced flow is called streaming. Acoustic cavitation is the action of acoustic fields within a fluid to generate bubbles and cause volume pulsation or even collapse in response to the acoustic field. The result can be heat generation and associated free radical generation, microstreaming of fluid around the bubble, radiation forces generated by the scattered acoustic field from the bubble, and mechanical actions from bubble collapse. The interaction of acoustic fields with bubbles or “gas bodies” (as they are generally called) has been a significant area of bioeffects research in recent years.

THERMAL EFFECTS Ultrasound Produces Heat As ultrasound propagates through the body, energy is lost through attenuation. Attenuation causes loss of penetration and the inability to image deeper tissues. Attenuation is the result of two processes, scattering and absorption. Scattering of the ultrasound results from the redirection of the acoustic energy by tissue encountered during propagation. With diagnostic ultrasound, some of the acoustic energy transmitted into the tissue is scattered back in the direction of the transducer, termed backscatter, which allows a signal to be detected and images made. Energy also is lost along the propagation

The rate of temperature increase in tissues exposed to ultrasound depends on a several factors, including spatial focusing, ultrasound frequency, exposure duration, and tissue type.

Spatial Focusing Ultrasound systems use multiple techniques to concentrate or focus ultrasound energy and improve the quality of measured signals. The analogy for light is that of a magnifying glass. The glass collects all the light striking its surface and concentrates it into a small region. In sonography and acoustics in general, the term intensity is used to describe the spatial distribution of ultrasonic power (energy per unit time), where intensity = power/ area and the area refers to the cross-sectional area of the ultrasound beam. Another common beam dimension is the beam width at a specified location of the field. If the same ultrasonic power is concentrated into a smaller area, the intensity will increase. Focusing in an ultrasound system can be used to improve the spatial resolution of the images. The side effect is an increased potential for bioeffects caused by heating and cavitation. In general, the greatest heating potential is between the scanhead and the focus, but the exact position depends on the focal distance, tissue properties, and heat generated within the scanhead itself. Returning to the magnifying glass analogy, most children learn that the secret to incineration is a steady hand. Movement distributes the power of the light beam over a larger area, thereby reducing its intensity. The same is true in ultrasound imaging. Thus, imaging systems that scan a beam through tissue reduce the spatial average intensity. Spectral Doppler and M-mode ultrasound imaging maintain the ultrasound beam in a stationary position (both considered unscanned modes) and therefore provide no opportunity to distribute the ultrasonic power spatially, whereas color flow Doppler, power mode Doppler, and B-mode (often called gray-scale) ultrasound imaging require that the beam be moved to new locations (scanned modes) at a rate sufficient to produce the real-time nature of these imaging modes.

Temporal Considerations The ultrasound power is the temporal rate at which ultrasound energy is produced. Therefore, controlling how ultrasound is produced in time seems a reasonable method for limiting its effects.

36   PART I  ■  Physics TISSUE ATTENUATION

p+ Attenuation (dB/cm/MHz)

25

p–

TP

10 5

TA

Time

FIGURE 2-1.  Pressure and intensity parameters measured in medical ultrasound. The variables are defined as follows: p+, peak positive pressure in waveform; p−, peak negative pressure in waveform; TP, temporal peak; PA, pulse average; and TA, temporal average.

Ultrasound can be produced in bursts rather than continuously. Ultrasound imaging systems operate on the principle of pulse-echo, in which a burst of ultrasound is emitted, followed by a quiescent period listening for echoes to return. This pulsed wave ultrasound is swept through the image plane numerous times during an imaging sequence. On the other hand, ultrasound may be transmitted in a continuous wave (CW) mode, in which the ultrasound transmission is not interrupted. The temporal peak intensity refers to the largest intensity at any time during ultrasound exposure (Fig. 2-1). The pulse average intensity is the average value over the ultrasound pulse. The temporal average is the average over the entire pulse repetition period (elapsed time between onset of ultrasound bursts). The duty factor is defined as the fraction of time the ultrasound field is “on.” With significant time “off ” between pulses (small duty factor), the temporal average value will be significantly smaller. For example, a duty factor of 10% will reduce the temporal average intensity by a factor of 10 compared to the pulse average. The time-averaged quantities are the variables most related to the potential for thermal bioeffects. Combining temporal and spatial information results in common terms such as the spatial peak, temporal average intensity (ISPTA) and spatial average, temporal average intensity (ISATA). The overall duration, or dwell time, of the ultrasound exposure to a particular tissue is important because longer exposure of the tissue may increase the risk of bioeffects. The motion of the scanhead during an examination reduces the dwell time within a particular region of the body and can minimize the potential for bioeffects

Sk ar in til ag In fa nt e sk ul l Sk ul l

0

ni o flu tic id Bl oo d Br ai n Li ve M r us cl e Fa Te t nd on

PA

15

Am

Instantaneous intensity

Pulse length (temporal duration)

20

C

Pressure

Pulse repetition period

Tissue Type

FIGURE 2-2.  Tissue attenuation. Values for types of human tissue at body temperature. (Data from Duck FA, Starritt HC, Anderson SP. A survey of the acoustic output of ultrasonic Doppler equipment. Clin Phys Physiol Meas 1987;8:39-49.)

of ultrasound. Therefore, performing an efficient scan, spending only the time required for diagnosis, is a simple way to reduce exposure.

Tissue Type Numerous physical and biologic parameters control heating of tissues. Absorption is normally the dominant contributor to attenuation in soft tissue. The attenuation coefficient is the attenuation per unit length of sound travel and is usually given in decibels per centimeters-megahertz (dB/cm-MHz). The attenuation typically increases with increasing ultrasound frequency. The attenuation ranges from a negligible amount for fluids (e.g., amniotic fluid, blood, urine) to the highest value for bone, with some variation among different soft tissue types (Fig. 2-2). Another important factor is the body’s ability to cool tissue through blood perfusion. Well-perfused tissue will more effectively regulate its temperature by carrying away the excess heat produced by ultrasound. The exception is when heat is deposited too rapidly, as in therapeutic thermal ablation.1 Bone and soft tissue are two specific areas of interest based on the differences in heating phenomena. Bone has high attenuation of incident acoustic energy. In examinations during pregnancy, calcified bone is typically subjected to ultrasound, as in measurement of the biparietal diameter (BPD) of the skull. Fetal bone contains increasing degrees of mineralization as gestation progresses, thereby increasing risk of localized heating. Special heating situations relevant to obstetric ultrasound examinations may also occur in soft tissue, where

Chapter 2  ■  Biologic Effects and Safety   37

overlying structures provide little attenuation of the field, such as the fluid-filled amniotic sac.

Bone Heating The absorption of ultrasound at bone allows for rapid deposition of energy from the field into a limited volume of tissue. The result can be a significant temperature rise. For example, Carstensen et al.2 combined an analytic approach and experimental measurements of the temperature rise in mouse skull exposed to CW ultrasound to estimate the temperature increments in bone exposures. Because bone has a large absorption coefficient, the incident ultrasonic energy is assumed to be absorbed in a thin planar sheet at the bone surface. The temperature rise of mouse skull has been studied in a 3.6-MHz focused beam with a beam width of 2.75 mm (Fig. 2-3). The temporal average intensity in the focal region was 1.5 W/cm2. One of two models (upper curve in Fig. 2-3) in common use3 predicts values for the temperature rise about 20% greater than that actually measured in this experiment.1 Thus the theoretical model is conservative in nature. Similarly for the fetal femur, Drewniak et al.4 indicated that the size and calcification state of the bone contributed to the ex vivo heating of bone (Table 2-1). To put this in perspective and to illustrate the operator’s role in controlling potential heating, consider the following scenario. By reducing the output power of an ultrasound scanner by 10 dB, the predicted temperature rise

8

Temperature increment (°C)

7 6 5 4

would be reduced by a factor of 10, making the increase of 3° C seen by these researchers (Table 2-1) virtually nonexistent. This strongly suggests the use of maximum gain and reduction in output power during ultrasound examinations (see later section on controlling ultrasound output). In fetal examinations an attempt should be made to maximize amplifier gain because this comes at no cost to the patient in terms of exposure. Distinctions are often made between bone positioned deep to the skin at the focal plane of the transducer and bone near the skin surface, as when considering transcranial applications. This distinction is discussed later with regard to the thermal index.

Soft Tissue Heating Two clinical situations for ultrasound exposure in soft tissue are particularly relevant to obstetric/gynecologic applications. First, a common scenario involves scanning through a full bladder. The urine is a fluid with a relatively low ultrasound attenuation coefficient. The reduced attenuation allows larger acoustic amplitudes to be applied deeper within the body. Second, the propagating wave may experience finite amplitude distortion, resulting in energy being shifted by a nonlinear process from lower to higher frequencies. The result is a shockwave where a gradual wave steepening results in a waveform composed of higher-frequency components (Fig. 2-4). Attenuation increases with increasing frequency; therefore the absorption of a large portion of the energy in such a wave occurs over a much shorter distance, concentrating the energy deposition in the first tissue encountered, which may include the fetus. Ultrasound imaging systems now include specific modalities that rely on nonlinear effects. In tissue harmonic imaging, or native harmonic imaging, the image is created using the backscatter of harmonic components induced by nonlinear propagation of the ultrasound field. This has distinct advantages in terms of reducing image artifacts and improving lateral resolution in particular. In these nonlinear imaging modes the acoustic

3 2

TABLE 2-1.  FETAL FEMUR TEMPERATURE INCREMENTS* AT 1 W/cm2

1 0 0.0

0.5

1.0

1.5

Exposure time (minutes)

FIGURE 2-3.  Heating of mouse skull in a focused sound field. For these experiments, frequency was 3.6 MHz, and temporal average focal intensity was 1.5 W/cm2. Solid circles: Young (6 mo) mice (N = 4); vertical bars: two standard errors in height; top curves: theoretical estimation of the temperature increases by Nyborg.3 (From Carstensen EL, Child SZ, Norton S, et al. Ultrasonic heating of the skull. J Acoust Soc Am 1990;87:1310-1317.)

Gestational Age (days)

Diameter (mm)

Temperature Increments (° C)

59 78 108

0.5 1.2 3.3

0.10 0.69 2.92

From Drewniak JL, Carnes KI, Dunn F. In vitro ultrasonic heating of fetal bone. J Acoust Soc Am 1989;86:1254-1258. *Temperature increments in human fetal femur exposed for 20 seconds were found to be approximately proportional to incident intensity.

38   PART I  ■  Physics 58 55 Temperature (°C)

Pressure

+

Shocked Normal

52 49 46 43

|

40 37 0.1

1

Time

FIGURE 2-4.  Effect of finite amplitude distortion on a propagating ultrasound pulse. Note the increasing steepness in the pulse, which contains higher-frequency components.

output must be sufficiently high to produce the effect. The acoustic power currently used is still within the FDA limits, but improvements in image quality using such modes may create the need to modify or relax the regulatory restrictions. Transvaginal ultrasound is important to note because of the proximity of the transducer to sensitive tissues such as the ovaries. As discussed later, temperature increases near the transducer may provide a heat source at sites other than the focus of the transducer. In addition, the transducer face itself may be a significant heat source because of inefficiencies in its conversion of electric to acoustic energy. Therefore, such factors must be considered in the estimation of potential thermal effects in transvaginal ultrasound and other endocavitary applications.

Hyperthermia and Ultrasound Safety Knowledge of the bioeffects for ultrasound heating is based on the experience available from other, more common forms of hyperthermia, which serve as a basis for safety criteria. Extensive data exist on the effects of short-term and extended temperature increases, or hyperthermia. Teratogenic effects from hyperthermia have been demonstrated in birds, all the common laboratory animals, farm animals, and nonhuman primates.5 The wide range of observed bioeffects, from subcellular chemical alterations to gross congenital abnormalities and fetal death, is an indication of the effectiveness or universality of hyperthermic conditions for perturbing living systems.6 The National Council on Radiation Protection and Measurements (NCRP) Scientific Committee on Biological Effects of Ultrasound compiled a comprehensive list of the lowest reported thermal exposures producing teratogenic effects.7,8 Examination of these data indicated a lower boundary for observed thermally

10

100

1000

Time(s)

FIGURE 2-5.  Conservative boundary curve for nonfetal bioeffects caused by a thermal mechanism. Note the increase in temperature tolerance associated with shorter durations of exposures, a modification to the earlier AIUM Conclusions Regarding Heat statement (March 26, 1997). AIUM approved a revised thermal statement on April 6, 2009. For a complete description of the origins of this curve, see O’Brien et al.10 (From O’Brien WD Jr, Deng CX, Harris GR, et al. The risk of exposure to diagnostic ultrasound in postnatal subjects: thermal effects. J Ultrasound Med 2008;27:517-535.)

induced bioeffects. Questions remain, however, about the relevance of this analysis of hyperthermia to the application of diagnostic ultrasound.9 More recently, after a careful literature review, O’Brien et al.10 suggested a more detailed consideration of thermal effects with regard to short-duration exposures. Figure 2-5 shows the recommended approach to addressing the combination of temperature and duration of exposure. Note that the tolerance of shorter durations and higher temperatures suggests a substantial safety margin for diagnostic ultrasound. Regardless, it is beneficial to provide feedback to the ultrasound operator as to the relative potential for a temperature rise in a given acoustic field under conditions associated with a particular examination. This will allow an informed decision as to the exposure needed to obtain diagnostically relevant information.

Thermal Index Based on analysis of hyperthermia data, NCRP proposed a general statement concerning the safety of ultrasound examinations in which no temperature rise greater than 1° C is expected. In an afebrile patient within this limit, NCRP concluded that there was no basis for expecting an adverse effect. In cases where the temperature rise might be greater, the operator should weigh the benefit against the potential risk. To assist in this decision, given the range of different imaging conditions seen in practice, a thermal index (TI) was approved as part of the Standard for Real-Time Display of Thermal and Mechanical Acoustical Output Indices on Diagnostic Ultrasound Equipment of the American Institute of Ultrasound in Medicine (AIUM).11 This standard

Chapter 2  ■  Biologic Effects and Safety   39

provides the operator with an indication of the relative potential risk of heating tissue, with calculations based on the imaging conditions and an on-screen display showing the TI.

THE THERMAL INDEX To more easily inform the physician of the operating conditions that could, in some cases, lead to a temperature elevation of 1°C, a thermal index is defined as TI =

W0 Wdeg

where Wdeg is the ultrasonic source power (in watts) calculated as capable of producing a 1°C temperature elevation under specific conditions. W0 is the ultrasonic source power (in watts) being used during the current exam. Reproduced with permission of American Institute of Ultrasound in Medicine (AIUM).

The NCRP ultrasound committee introduced the TI concept.7 The purpose of the TI is to provide an indication of the relative potential for increasing tissue temperature, but it is not meant to provide the actual temperature rise. The NCRP recommended two tissue models to aid in the calculation of the ultrasound power that could raise the temperature in tissue by 1° C: (1) a homogeneous model in which the attenuation coefficient is uniform throughout the region of interest, and (2) a fixed-attenuation model in which the minimum attenuation along the path from transducer to a distant anatomic structure is independent of the distance because of a low-attenuation fluid path (e.g., amniotic fluid).7,12,13 Because of concern for the patient, it was recommended that “reasonable worst case” assumptions be made with respect to estimation of temperature elevations in vivo. The FDA, AIUM, and National Electrical Manufacturers Association (NEMA) adopted the TI as part of the output display standard. They advocate estimating the effect of attenuation in the body by reducing the acoustic power/output of the scanner (W0) by a derating factor equal to 0.3 dB/cm-MHz for the soft tissue model.11 The AIUM Thermal Index Working Group considered three tissue models: (1) the homogeneous tissue or soft tissue model, (2) a tissue model with bone at the focus, and (3) a tissue model with bone at the surface, or transcranial model.11 The TI takes on three different forms for these tissue models.

Homogeneous Tissue Model (Soft Tissue) The assumption of homogeneity helps simplify the determination of the effects of acoustic propagation and

attenuation, as well as the heat transfer characteristics of the tissue. Providing one of the most common applications for ultrasound imaging, this model applies to situations where bone is not present and can generally be used for fetal examinations during the first trimester (low calcification in bone). In the estimation of potential heating, many assumptions and compromises had to be made to calculate a single quantity that would guide the operator. Calculations of the temperature rise along the axis of a focused beam for a simple, spherically curved, single-element transducer result in two thermal peaks. The first is in the near field (between the transducer and the focus), and the second appears close to the focal region.14,15 The first thermal peak occurs in a region with low ultrasound intensity and wide beam width. When the beam width is large, cooling will occur mainly because of perfusion. In the near field the magnitude of the local intensity is the chief determinant of the degree of heating. The second thermal peak occurs at the location of high intensity and narrow beam width at or near the focal plane. Here the cooling is dominated by conduction, and the total acoustic power is the chief determinant of the degree of heating. Given the thermal “twin peaks” dilemma, the AIUM Thermal Index Working Group compromised in creating a TI that included contributions from both heating domains.11 Their rationale was based on the need to minimize the acoustic measurement load for manufacturers of ultrasound systems. In addition, adjustments had to be made to compensate for effects of the large range of potential apertures. The result is a complicated series of calculations and measurements that must be performed, and to the credit of the many manufacturers, there has been considerable effort in implementing a display standard to provide user feedback. Different approaches to these calculations are being considered,10 but changes will require that the currently accepted implementation be reexamined and approved for use by the FDA and considered by the International Electrotechnical Commission (IEC), a standards organization.

Tissue Model with Bone at the Focus (Fetal Applications) Applications of ultrasound in which the acoustic beam travels through soft tissue for a fixed distance and impinges on bone occur most often in obstetric scanning during the second and third trimesters. Carson et al.13 recorded sonographic measurements of the maternal abdominal wall thickness in various stages of pregnancy. Based on their results, the NCRP recommended that the attenuation coefficients for the first, second, and third trimesters be 1.0, 0.75, and 0.5 dB/MHz, respectively.7 These values represent “worst case” estimates. In addition, Siddiqi et al.16 determined the average tissue attenuation coefficient for transabdominal insonification (exposure to ultrasound waves) in a patient population

40   PART I  ■  Physics

of nonpregnant, healthy volunteers was 2.98 dB/MHz. This value represents an average measured value and is much different than the worst-case estimates previously listed. This leads to considerable debate on how such parameters should be included in an index. In addition, bone is a complex, hard connective tissue with a calcified collagenous intercellular substance. Its absorption coefficient for longitudinal waves is a factor of 10 greater than that for most soft tissues (see Fig. 2-2). Shear waves are also created in bone as sound waves strike bone at oblique incidence. The absorption coefficients for shear waves are even greater than those for longitudinal waves.17-19 Based on the data of Carstensen et al.2 described earlier, the NCRP proposed a thermal model for bone heating. Using this model, the thermal index for bone (TIB) is estimated for conditions in which the focus of the beam is at or near bone. Again, assumptions and compromises had to be made to develop a functional TI for the case of bone exposure, as follows: • For unscanned mode transducers (operating in a fixed position) with bone in the focal region, the location of the maximum temperature increase is at the surface of the bone. Therefore the TIB is calculated at an axial distance where it is maximized, a worst-case assumption. • For scanned modes, the thermal index for soft tissue (TIS) is used because the temperature increase at the surface is either greater than or approximately equal to the temperature increase with bone in the focus.

Tissue Model with Bone at the Surface (Transcranial Applications) For adult cranial applications, the same model as that with bone at the focus is used to estimate the temperature distribution in situ. However, because the bone is located at the surface, immediately after the acoustic beam enters the body, attenuation of the acoustic power output is not included.11 In this situation the equivalent beam diameter at the surface is used to calculate the acoustic power.

Estimate of Thermal Effects Ultrasound users should keep in mind several points when referring to the thermal index as a means of estimating the potential for thermal effects. First, the TI is not synonymous with temperature rise. A TI equal to 1 does not mean the temperature will rise 1° C. An increased potential for thermal effects can be expected as TI increases. Second, a high TI does not mean that bioeffects are occurring, but only that the potential exists. Factors that may reduce the actual temperature rise may not be considered by the thermal models employed for TI calculation. However, TI should be

monitored during examinations and minimized when possible. Finally, there is no consideration in the TI for the duration of the scan, so minimizing the overall examination time will reduce the potential for effects.

Summary Statement on Thermal Effects The AIUM statement concerning thermal effects of ultrasound includes several conclusions that can be summarized as follows20: • Adult examinations resulting in a temperature rise of up to 2° C are not expected to cause bioeffects. (Many ultrasound examinations fall within these parameters.) • A significant number of factors control heat production by diagnostic ultrasound. • Ossified bone is a particularly important concern for ultrasound exposure. • A labeling standard now provides information concerning potential heating in soft tissue and bone. • Even though an FDA limit exists for fetal exposures, predicted temperature rises can exceed 2° C. • Thermal indices are expected to track temperature increases better than any single ultrasonic field parameter.

EFFECTS OF ACOUSTIC CAVITATION Potential Sources for Bioeffects Knowledge concerning the interaction of ultrasound with gas bodies (which many term “cavitation”) has significantly increased recently, although it is not as extensive as that for ultrasound thermal effects and other sources of hyperthermia. Acoustic cavitation inception is demarcated by a specific threshold value: the minimum acoustic pressure necessary to initiate the growth of a cavity in a fluid during the rarefaction phase of the cycle. Several parameters affect this threshold, including initial bubble or cavitation nucleus size, acoustic pulse characteristics (e.g., center frequency, pulse repetition frequency, pulse duration), ambient hydrostatic pressure, and host fluid parameters (e.g., density, viscosity, compressibility, heat conductivity, surface tension). Inertial cavitation refers to bubbles that undergo large variations from their equilibrium sizes in a few acoustic cycles. Specifically during contraction, the surrounding fluid inertia controls the bubble motion.21 Large acoustic pressures are necessary to generate inertial cavitation, and the collapse of these cavities is often violent. The effect of preexisting cavitation nuclei may be one of the principal controlling factors in mechanical effects that result in biologic effects. The body is such an excellent filter that these nucleation sites may be found

AIUM STATEMENT ON HEAT—THERMAL BIOEFFECTS Approved April 6, 2009 1. Excessive temperature increase can result in toxic effects in mammalian systems. The biological effects observed depend on many factors, such as the exposure duration, the type of tissue exposed, its cellular proliferation rate, and its potential for regeneration. Age and stage of development are important factors when considering fetal and neonatal safety. Temperature increases of several degrees Celsius above the normal core range can occur naturally. The probability of an adverse biological effect increases with the duration of the temperature rise. 2. In general, adult tissues are more tolerant of temperature increases than fetal and neonatal tissues. Therefore, higher temperatures and/or longer exposure durations would be required for thermal damage. The considerable data available on the thermal sensitivity of adult tissues support the following inferences: For exposure durations up to 50 hours, there have been no significant, adverse biological effects observed due to temperature increases less than or equal to 2°C above normal. For temperature increases between 2°C and 6°C above normal, there have been no significant, adverse biological effects observed due to temperature increases less than or equal to 6 − log10(t/60)/0.6 where t is the exposure duration in seconds. For example, for temperature increases of 4°C and 6°C, the corresponding limits for the exposure durations t are 16 min and 1 min, respectively. For temperature increases greater than 6°C above normal, there have been no significant, adverse biological effects observed due to temperature increases less than or equal to 6 − log10(t/60)/0.3 where t is the exposure duration in seconds. For example, for temperature increases of 9.6°C and 6.0°C, the corresponding limits for the exposure durations t are 5 and 60 seconds, respectively. For exposure durations less than 5 seconds, there have been no significant, adverse biological effects observed due to temperature increases less than or equal to 9 − log10(t/60)/0.3 where t is the exposure duration in seconds. For example, for temperature increases of 18.3°C, 14.9°C, and 12.6°C, the corresponding limits for the exposure durations t are 0.1, 1, and 5 seconds, respectively. 3. Acoustic output from diagnostic ultrasound devices is sufficient to cause temperature elevations in fetal tissue. Although fewer data are available for fetal tissues, the following conclusions are justified: In general, temperature elevations become progressively greater from B-mode to color Doppler to spectral Doppler applications. For identical exposure conditions, the potential for thermal bioeffects increases with the dwell time during examination. For identical exposure conditions, the temperature rise near bone is significantly greater than in soft tissues, and it increases with ossification

development throughout gestation. For this reason, conditions where an acoustic beam impinges on ossifying fetal bone deserve special attention due to its close proximity to other developing tissues. The current FDA regulatory limit for ISPTA.3 is 720 mW/cm2. For this, and lesser intensities, the theoretical estimate of the maximum temperature increase in the conceptus can exceed 2°C. Although, in general, an adverse fetal outcome is possible at any time during gestation, most severe and detectable effects of thermal exposure in animals have been observed during the period of organogenesis. For this reason, exposures during the first trimester should be restricted to the lowest outputs consistent with obtaining the necessary diagnostic information. Ultrasound exposures that elevate fetal temperature by 4°C above normal for 5 minutes or more have the potential to induce severe developmental defects. Thermally induced congenital anomalies have been observed in a large variety of animal species. In current clinical practice, using commercially available equipment, it is unlikely that such thermal exposure would occur at a specific fetal anatomic site. Transducer self-heating is a significant component of the temperature rise of tissues close to the transducer. This may be of significance in transvaginal scanning, but no data for the fetal temperature rise are available. 4. The temperature increase during exposure of tissues to diagnostic ultrasound fields is dependent upon (a) output characteristics of the acoustic source such as frequency, source dimensions, scan rate, power, pulse repetition frequency, pulse duration, transducer self-heating, exposure time, and wave shape and (b) tissue properties such as attenuation, absorption, speed of sound, acoustic impedance, perfusion, thermal conductivity, thermal diffusivity, anatomical structure, and nonlinearity parameter. 5. Calculations of the maximum temperature increase resulting from ultrasound exposure in vivo are not exact because of the uncertainties and approximations associated with the thermal, acoustic, and structural characteristics of the tissues involved. However, experimental evidence shows that calculations are generally capable of predicting measured values within a factor of two. Thus, such calculations are used to obtain safety guidelines for clinical exposures where direct temperature measurements are not feasible. These guidelines, called thermal indices,* provide a real-time display of the relative probability that a diagnostic system could induce thermal injury in the exposed subject. Under most clinically relevant conditions, the soft tissue thermal index, TIS, and the bone thermal index, TIB, either overestimate or closely approximate the best available estimate of the maximum temperature increase (ΔTmax). For example, if TIS = 2, then ΔTmax ≤ 2°C.

Reprinted with permission of AIUM. *Thermal indices are the nondimensional ratios of the estimated temperature increases to 1°C for specific tissue models. See American Institute of Ultrasound in Medicine. Standard for real-time display of thermal and mechanical acoustic output indices on diagnostic ultrasound equipment, Revision 2, Rockville, Md, 2001, AIUM and National Electrical Manufacturers Association.

42   PART I  ■  Physics

FIGURE 2-6.  Acoustic cavitation bubbles. This cavitation activity is being generated in water using a common therapeutic ultrasound device. (Courtesy National Center for Physical Acoustics, University of Mississippi.)

only in small numbers and at selected sites. For example, if water is filtered down to 2 µm, the cavitation threshold doubles.22 Theoretically, the tensile strength of water that is devoid of cavitation nuclei is about 100 megapascals (MPa).23 Various models have been suggested to explain bubble formation in animals,24,25 and these models have been used extensively in cavitation threshold determination. One model is used in the prediction of SCUBA diving tables and may also have applicability to patients.26 It remains to be seen how well such models will predict the nucleation of bubbles from diagnostic ultrasound in the body. Figure 2-6 shows a 1-MHz therapeutic ultrasound unit generating bubbles in gas-saturated water. The particular medium and ultrasound parameters were chosen to optimize the conditions for cavitation. Using continuous wave ultrasound and many preexisting gas pockets in the water set the stage for the production of cavitation. Even though these acoustic pulses are longer than those typically used in diagnostic ultrasound, cavitation effects have also been observed with diagnostic pulses in fluids.27 Ultrasound contrast agents composed of stabilized gas bubbles should provide a source of cavitation nuclei, as discussed later.

Sonochemistry Free radical generation and detection provide a means to observe cavitation and to gauge its strength and potential for damage. The sonochemistry of free radicals is the result of very high temperatures and pressures within the rapidly collapsing bubble. These conditions can even generate light, or sonoluminescence.28 With the addition of the correct compounds, chemical luminescence can also be used for free radical detection and can be generated with short pulses similar to that used in diagnostic ultrasound.29 Figure 2-7 shows chemiluminescence generated by a therapeutic ultrasound device; the

FIGURE 2-7.  Chemical reaction induced by cavitation producing visible light. The reaction is the result of free radical production. (Courtesy National Center for Physical Acoustics, University of Mississippi.)

setup is backlighted (in red) to show the bubbles and experimental apparatus. The chemiluminescence emissions are the blue bands seen through the middle of the liquid sample holder. The light emitted is sufficient to be seen by simply adapting one’s eyes to darkness. Electron spin resonance can also be used with molecules that trap free radicals to detect cavitation activity capable of free radical production.30 A number of other chemical detection schemes are presently employed to detect cavitation from diagnostic devices in vitro.

Evidence of Cavitation from Lithotripters It is possible to generate bubbles in vivo using short pulses with high amplitudes of an extracorporeal shockwave lithotripter (ESWL). The peak positive pressure for lithotripsy pulses can be as high as 50 MPa, with the negative pressure about 20 MPa. Finite amplitude distortion causes high frequencies to appear in highamplitude ultrasound fields. Although ESWL pulses have significant energy at high frequencies because of finite amplitude distortion, a large portion of the energy is actually in the 100-kHz range, much lower than frequencies in diagnostic scanners. The lower frequency makes cavitation more likely. Aymé and Carstensen31 showed that the higher-frequency components in nonlinearly distorted pulses contribute little to the killing of Drosophila larvae. Interestingly, increasing evidence indicates that collapsing bubbles play a role in stone disruption.32-34 A bubble collapsing near a surface may form a liquid jet through its center, which strikes the surface (Fig. 2-8). Placing a sheet of aluminum foil at the focus of a lithotripter generates small pinholes.32 The impact is even sufficient to pit solid brass and aluminum plates.

Chapter 2  ■  Biologic Effects and Safety   43

FIGURE 2-8.  Collapsing bubble near a boundary. When cavitation is produced near boundaries, a liquid jet may form through the center of a bubble and strike the boundary surface. (Courtesy Lawrence A. Crum.)

Clearly, lithotripsy and diagnostic ultrasound differ in the acoustic power generated and are not comparable in the bioeffects produced. However, some diagnostic devices produce peak rarefactional pressures greater than 3 MPa, which is in the lower range of lithotripter outputs.35-37 Lung damage and surface petechiae have been noted as side effects of ESWL in clinical cases.38 Inertial cavitation was suspected as the cause, prompting several researchers to study the effects of diagnostic ultrasound exposure on the lung parenchyma.39,40

Bioeffects in Lung and Intestine Lung tissue and intestinal tissue are key locations for examining for bioeffects of diagnostic ultrasound.39 The presence of air in the alveolar spaces constitutes a significant source of gas bodies. Child et al.40 measured threshold pressures for hemorrhage in mouse lung exposed to 1- to 4-MHz short-pulse diagnostic ultrasound (i.e., 10and 1-mm pulse durations). The threshold of damage in murine lung at these frequencies was 1.4 MPa. Pathologic features of this damage included extravasation of blood cells into the alveolar spaces.41 The authors hypothesized that cavitation, originating from gas-filled alveoli, was responsible for the damage. Their data provided the first direct evidence that clinically relevant, pulsed ultrasound exposures produce deleterious effects in mammalian tissue in the absence of significant heating. Hemorrhagic foci induced by 4-MHz pulsed Doppler ultrasound have also been reported in the monkey.42 Damage in the monkey lung was of a significantly lesser degree than that in the mouse. In these studies it was impossible to show categorically that bubbles induced these effects because the cavitation-induced bubbles were not observed. Thresholds for petechial hemorrhage in the lung caused by ultrasound have been measured in mouse, rat, rabbit and pig.43-45 Direct mechanical stresses associated with

propagation of ultrasound in the lung were believed to contribute to the damage observed.39,46 Thresholds for hemorrhage in the murine intestine exposed to pulsed ultrasound have also been determined.47 Kramer et al.48 assessed cardiopulmonary function in rats exposed to pulsed ultrasound well above the acoustic output threshold of damage, at a mechanical index (MI) of 9.7 (see later discussion). Measurements of cardiopulmonary function included arterial blood pressure, heart rate, respiratory rate, and arterial blood gases (Pco2 and Po2). If only one side of the rat lung was exposed, the cardiopulmonary measurements did not change significantly between baseline and postexposure values because of the functional respiratory reserve in the unexposed lobes. However, when both sides of the lung had significant ultrasound-induced lesions, the rats were unable to maintain systemic arterial pressure or resting levels of arterial Po2. Further studies are required to determine the relevance of these findings to humans. In general, tissues containing air (or stabilized gas) are more susceptible to damage than those tissues without gas. Also, no confirmed reports of petechial hemorrhage have been noted in animal studies below an MI of 0.4.

Ultrasound Contrast Agents The apparent absence of cavitation in many locations in the body can result from the lack of available cavitation nuclei. Based on evidence in the lung and intestine in mammalian models described earlier, the presence of gas bodies clearly reduces the requisite acoustic field for producing bioeffects. Many ultrasound contrast agents are composed of stabilized gas bubbles, so they could provide readily available nuclei for potential cavitation activity. This makes the investigation of bioeffects in the presence of ultrasound contrast agents an important area of research.49-51 Studies have also shown that ultrasound exposure in the presence of contrast agents produces small vascular petechiae and endothelial damage in mammalian systems.52-57 Acoustic emissions from activated microbubbles correlate with the degree of vascular damage.54,55 As a result, the AIUM has updated a safety statement on the bioeffects of diagnostic ultrasound with gas body contrast agents. This bioeffect may occur, but the issue remains whether it constitutes a significant physiologic risk. The safety statement is designed to make sonographers and physicians aware of the potential for bioeffects in the presence of gas contrast agents and allow them to make an informed decision based on a risk/benefit assessment. Some research also indicates the production of premature ventricular contractions (PVCs) during cardiac scanning in the presence of ultrasound contrast agents. At least one human study indicated an increase in PVCs only when ultrasound imaging was performed with a

44   PART I  ■  Physics

AIUM STATEMENT ON BIOEFFECTS OF DIAGNOSTIC ULTRASOUND WITH GAS BODY CONTRAST AGENTS Approved November 8, 2008 Presently available ultrasound contrast agents consist of suspensions of gas bodies (stabilized gaseous microbubbles). The gas bodies have the correct size for strong echogenicity with diagnostic ultrasound and also for passage through the microcirculation. Commercial agents undergo rigorous clinical testing for safety and efficacy before Food and Drug Administration approval is granted, and they have been in clinical use in the United States since 1994. Detailed information on the composition and use of these agents is included in the package inserts. To date, diagnostic benefit has been proven in patients with suboptimal echocardiograms to opacify the left ventricular chamber and to improve the delineation of the left ventricular endocardial border. Many other diagnostic applications are under development or clinical testing. Contrast agents carry some potential for nonthermal bioeffects when ultrasound interacts with the gas bodies. The mechanism for such effects is related to the physical phenomenon of acoustic cavitation. Several published reports describe adverse bioeffects in mammalian tissue in vivo resulting from exposure to diagnostic ultrasound with gas body contrast agents in the circulation. Induction of premature ventricular contractions by triggered contrast echocardiography in humans has been reported for a noncommercial agent and in laboratory animals for commercial agents. Microvascular leakage, killing of cardiomyocytes, and glomerular capillary hemorrhage, among other bioeffects, have been reported in animal studies. Two medical ultrasound societies have examined this

potential risk of bioeffects in diagnostic ultrasound with contrast agents and provide extensive reviews of the topic: the World Federation for Ultrasound in Medicine and Biology (WFUMB) Contrast Agent Safety Symposium* and the American Institute of Ultrasound in Medicine 2005 Bioeffects Consensus Conference.49 Based on review of these reports and of recent literature, the Bioeffects Committee issues the following statement: Induction of premature ventricular contractions, microvascular leakage with petechiae, glomerular capillary hemorrhage, and local cell killing in mammalian tissue in vivo have been reported and independently confirmed for diagnostic ultrasound exposure with a mechanical index (MI) above about 0.4 and a gas body contrast agent present in the circulation. Although the medical significance of such microscale bioeffects is uncertain, minimizing the potential for such effects represents prudent use of diagnostic ultrasound. In general, for imaging with contrast agents at an MI above 0.4, practitioners should use the minimal agent dose, MI, and examination time consistent with efficacious acquisition of diagnostic information. In addition, the echocardiogram should be monitored during high-MI contrast cardiac-gated perfusion echocardiography, particularly in patients with a history of myocardial infarction or unstable cardiovascular disease. Furthermore, physicians and sonographers should follow all guidance provided in the package inserts of these drugs, including precautions, warnings, and contraindications.

Reprinted with permission of AIUM. *Barnett SB. Safe use of ultrasound contrast agents. WFUMB Symposium on Safety of Ultrasound in Medicine: Ultrasound Contrast Agents. Ultrasound Med Biol 2007;33:171-172.

contrast agent, and not with ultrasound imaging alone or during injection of the agent without imaging.58 Another study59 revealed that oscillating microbubbles affect stretch activation channels60,61 in cardiac cells, which generates membrane depolarization and triggers action potentials and thus PVCs. The importance of this bioeffect is also being debated because there is a naturally occurring rate of PVCs, and a small increase may not be considered significant, particularly if the patient benefits from using the agent. Additional consideration might be given to patients with specific conditions in whom additional PVCs should be avoided. The consequences of the ultrasound contrast agent bioeffects reported thus far require more study. Although the potential exists for a bioeffect, its scale and influence on human physiology remain unclear. Contrast agents have demonstrated efficacy for specific indications, facilitating patient management.62 In addition, clinical trials

and marketing follow-up of many patients receiving ultrasound and contrast agents have reported few effects. In fact, recent evidence confirms the safety of ultrasound contrast agent use.63-66

Mechanical Index Calculations for cavitation prediction have yielded a trade-off between peak rarefactional pressure and frequency.67 This predicted trade-off assumes short-pulse (a few acoustic cycles) and low-duty cycle ultrasound ( 4 for extravasation of blood cells in mouse kidneys, and MI > 5.1 for hind limb paralysis in the mouse neonate. 6. For diagnostically relevant exposures (MI ≤ 1.9), no independently confirmed, biologically significant adverse nonthermal effects have been reported in mammalian tissues that do not contain well-defined gas bodies.

Reprinted with permission of AIUM. *The MI is equal to the derated peak rarefactional pressure (in MPa) at the point of the maximum derated pulse intensity integral divided by the square root of the ultrasonic center frequency (in MHz). See American Institute of Ultrasound in Medicine. Standard for real-time display of thermal and mechanical acoustic output indices on diagnostic ultrasound equipment, Revision 2. Rockville, Md, 2004, AIUM and National Electrical Manufacturers Association.

4. In the absence of gas bodies, the threshold for damage is much higher. (This is significant because ultrasound examinations may be performed predominantly in tissues with no identifiable gas bodies.)

OUTPUT DISPLAY STANDARD Several groups, including the FDA, AIUM, and NEMA, have developed the Standard for Real-Time Display of Thermal and Mechanical Acoustical Output Indices on Diagnostic Ultrasound Equipment, which introduces a method to provide the user with information concerning the thermal and mechanical indices. Real-time display of the MI and TI will allow a more informed decision on the potential for bioeffects during ultrasound examinations (Fig. 2-10). The standard requires dynamic updates

Ovarian follicles

FIGURE 2-10.  Display of bioeffects indices. Typical appearance of an ultrasound scanner display showing (right upper corner) the thermal index in bone (TIb) and mechanical index (MI) for an endocavitary transducer.

Chapter 2  ■  Biologic Effects and Safety   47

of the indices as instrument output is modified and allows the operator to learn how controls will affect these indices. Important points to remember about this display standard include the following: • The mechanical index should be clearly visible on the screen (or alert the operator by some other means) and should begin to appear when the instrument exceeds a value of 0.4. An exception is made for instruments incapable of exceeding index values of 1; these are not required to display the bioeffects indices such as MI. • Sometimes only one index (MI or TI) will be displayed at a time. The choice is often based on whether a given output condition is more likely to produce an effect by either mechanism. • The standard also requires that appropriate default output settings be in effect at power-up, new patient entry, or when changing to a fetal examination. After that time the operator can adjust the instrument output as necessary to acquire clinically useful information while attempting to minimize the index values. • As indicated previously, the bioeffects indices do not include any factors associated with the time taken to perform the scan. Efficient scanning is still an important component in limiting potential bioeffects. In the document Medical Ultrasound Safety, AIUM suggests that the operator ask the following four questions to use the output display effectively69: 1. Which index should be used for the examination being performed? 2. Are there factors present that might cause the reading to be too high or low? 3. Can the index value be reduced further even when it is already low? 4. How can the ultrasound exposure be minimized without compromising the scan’s diagnostic quality? Sonographers and physicians are being presented with real-time data on acoustic output of diagnostic scanners and are being asked not only to understand the manner in which ultrasound propagates through and interacts with tissue, but also to gauge the potential for adverse bioeffects. The output display is a tool that can be used to guide an ultrasound examination and control for potential adverse effects. The thermal and mechanical indices provide the user with more information and more responsibility in limiting output.

GENERAL AIUM SAFETY STATEMENTS It is important to consider some official positions concerning the status of bioeffects resulting from ultrasound. Most important is the high level of confidence in the safety of ultrasound in official statements. For example,

AIUM SAFETY STATEMENTS ON DIAGNOSTIC ULTRASOUND AIUM STATEMENT ON PRUDENT USE AND CLINICAL SAFETY

Approved October 1982; Revised and approved March 2007 Diagnostic ultrasound has been in use since the late 1950s. Given its known benefits and recognized efficacy for medical diagnosis, including use during human pregnancy, the American Institute of Ultrasound in Medicine herein addresses the clinical safety of such use: No independently confirmed adverse effects caused by exposure from present diagnostic ultrasound instruments have been reported in human patients in the absence of contrast agents. Biological effects (such as localized pulmonary bleeding) have been reported in mammalian systems at diagnostically relevant exposures, but the clinical significance of such effects is not yet known. Ultrasound should be used by qualified health professionals to provide medical benefit to the patient.

AIUM STATEMENT ON SAFETY IN TRAINING AND RESEARCH

Approved March 1983; Revised and approved March, 2007 Diagnostic ultrasound has been in use since the late 1950s. There are no confirmed adverse biological effects on patients resulting from this usage. Although no hazard has been identified that would preclude the prudent and conservative use of diagnostic ultrasound in education and research, experience from normal diagnostic practice may or may not be relevant to extended exposure times and altered exposure conditions. It is therefore considered appropriate to make the following recommendation: When examinations are carried out for purposes of training or research, the subject should be informed of the anticipated exposure conditions and how these compare with normal diagnostic practice. Reprinted with permission of AIUM.

in 2007 the AIUM reiterated its earlier statement concerning the clinical use of diagnostic ultrasound by stating that no known bioeffects have been confirmed with the use of present diagnostic equipment, and the patient benefits resulting from prudent use outweigh the risks, if any. Similarly, in commenting on the use of diagnostic ultrasound in research, by AIUM recommends that in the case of ultrasound exposure for other than direct medical benefit, the person should be informed concerning the exposure conditions and how these relate to normal exposures. For the most part, even examinations for research purposes are comparable to normal diagnostic exams and pose no additional risk. In

48   PART I  ■  Physics

fact, many research exams can be performed in conjunction with routine exams. The effects based on in vivo animal models can be summarized by the AIUM Statements on Heat— Thermal Bioeffects, Bioeffects of Diagnostic Ultrasound with Gas Body Contrast Agents, and Naturally Occurring Gas Bodies.20 No independently confirmed experimental evidence indicates damage in animal models below certain prescribed levels (temperature rises 10 mm Hg)86 (Fig. 4-31, C and D). • Paraumbilical vein: Runs in the falciform ligament and connects the left portal vein to the systemic epigastric veins near the umbilicus (CruveilhierBaumgarten syndrome)90 (Fig. 4-31, A). Some suggest that, if the hepatofugal flow in the patent paraumbilical vein exceeds the hepatopetal flow in the portal vein, patients may be protected from developing esophageal varices.91,92 • Splenorenal and gastrorenal: Tortuous veins may be seen in the region of the splenic and left renal

FIGURE 4-30.  Portal hypertension. Major sites of portosystemic venous collaterals. (From Subramanyam BR, Balthazar EJ, Madamba MR, et al: Sonography of portosystemic venous collaterals in portal hypertension. Radiology 1983;146:161-166.)

PORTOSYSTEMIC VENOUS COLLATERALS: MAJOR SITES IDENTIFIED ON ULTRASOUND 1. Gastroesophageal junction 2. Paraumbilical vein in falciform ligament 3. Splenorenal and gastrorenal veins 4. Intestinal-retroperitoneal anastomoses 5. Hemorrhoidal veins

hilus (Fig. 4-31, E and F), which represent collaterals between the splenic, coronary, and short gastric veins and the left adrenal or renal veins. • Intestinal: Regions in which the gastrointestinal tract becomes retroperitoneal so that the veins of the ascending and descending colon, duodenum, pancreas, and liver may anastomose with the renal, phrenic, and lumbar veins (systemic tributaries). • Hemorrhoidal: The perianal region where the superior rectal veins, which extend from the inferior mesenteric vein, anastomose with the systemic middle and inferior rectal veins. Duplex Doppler sonography provides additional information regarding direction of portal flow. False results may occur, however, when sampling is obtained from periportal collaterals in patients with portal vein thrombosis or hepatofugal portal flow.93 Normal portal venous flow rates will vary in the same individual, increasing postprandially and during inspiration83,94 and decreasing after exercise or in the upright position.95 An increase of less than 20% in the diameter of the portal

102   PART II  ■  Abdominal, Pelvic, and Thoracic Sonography

A

B

C

D

E

F

FIGURE 4-31.  Portal hypertension. A, Sagittal image of recanalized paraumbilical vein in patient with gross ascites. B, Sagittal image shows enlarged coronary vein running cephalad from the splenic vein (SV). C, Gray-scale image, and D, color Doppler image, show extensive varices in the distribution of the coronary vein. E, Gray-scale image, and F, color Doppler image, show splenic hilar varices.

vein with deep inspiration indicates portal hypertension with 81% sensitivity and 100% specificity.96 The normal portal vein demonstrates an undulating hepatopetal (toward the liver) flow. Mean portal venous flow velocity is approximately 15 to 18 cm/sec and varies with respiration and cardiac pulsation. As portal hyper-

tension develops, the flow in the portal vein loses its undulatory pattern and becomes monophasic. As the severity of portal hypertension increases, flow becomes biphasic and finally hepatofugal (away from the liver). Intrahepatic arterial-portal venous shunting may also be seen.

Chapter 4  ■  The Liver   103

Chronic liver disease is also associated with increased splanchnic blood flow. Recent evidence suggests that portal hypertension is partly caused by the hyperdynamic flow state of cirrhosis. Zweibel et al.97 found that blood flow was increased in the superior mesenteric arteries and splenic arteries of patients with cirrhosis and splenomegaly, compared with normal controls. Of interest, in patients with cirrhosis and normal-sized livers, splanchnic blood flow was not increased. Patients with isolated splenomegaly and normal livers were not included in this study. The limitations of Doppler sonography in the evaluation of portal hypertension include the inability to determine vascular pressures and flow rates accurately. Patients with portal hypertension are often ill, with contracted livers, abundant ascites, and floating bowel, all of which create a technical challenge. In a comparison of duplex Doppler sonography with MR angiography, MR imaging was superior in the assessment of patency of the portal vein and surgical shunts, as well as in detection of varices.98 However, when technically adequate, the Doppler study was accurate in the assessment of normal

portal anatomy and flow direction. Duplex Doppler sonography has the added advantages of decreased cost and portability of the equipment and therefore should be used as the initial screening method for portal hypertension.

Portal Vein Thrombosis Portal vein thrombosis has been associated with malignancy, including HCC, metastatic liver disease, carcinoma of the pancreas, and primary leiomyosarcoma of the portal vein,99 as well as with chronic pancreatitis, hepatitis, septicemia, trauma, splenectomy, portacaval shunts, hypercoagulable states such as pregnancy and in neonates, omphalitis, umbilical vein catheterization, and acute dehydration.100 Sonographic findings of portal vein thrombosis include echogenic thrombus within the lumen of the vein, portal vein collaterals, expansion of the caliber of the vein, and cavernous transformations100 (Figs. 4-32 and 4-33). Cavernous transformation of the portal vein refers to numerous wormlike vessels at the porta

A

B

C

D

FIGURE 4-32.  Portal vein thrombosis: benign and malignant. Malignant thrombus: transverse views of A, the vein at the porta hepatis, and B, left ascending left portal vein. Both are distended with occlusive thrombus. Benign thrombus: C, transverse, and D, sagittal, images of simple, bland nonocclusive thrombus in the left portal vein at the porta hepatis.

104   PART II  ■  Abdominal, Pelvic, and Thoracic Sonography

A

B

FIGURE 4-33.  Cavernous transformation of portal vein. A, Gray-scale image, and B, Color Doppler image. Numerous periportal collateral vessels are present.

A

B

FIGURE 4-34.  Metastasis to the portal vein from colon cancer. A, Sagittal view of the main portal vein at the porta hepatis, and B, subcostal oblique sonogram of the left ascending branch of the portal vein, show the portal vein is distended and highly echogenic (arrows). There is also evidence of cavernous transformation, an uncommon accompaniment of malignant portal vein occlusion.

hepatis, which represent periportal collateral circulation.101 This pattern is observed in long-standing thrombosis, requiring up to 12 months to occur, and thus is more likely to develop with benign disease.102 Acute thrombus may appear relatively anechoic and thus may be overlooked unless Doppler ultrasound interrogation is performed. Malignant thrombosis of the portal vein has a high association with HCC and is often expansive, as is malignant occlusion from other primary or secondary disease (Fig. 4-34). Doppler sonography is useful in distinguishing between benign and malignant portal vein thrombi in patients with cirrhosis. Both bland and malignant

thrombi may demonstrate continuous blood flow. Pulsatile flow, however, has been found to be 95% specific for the diagnosis of malignant portal vein thrombosis (see Fig. 4-32). The sensitivity was only 62% because many malignant thrombi are hypovascular.103

Budd-Chiari Syndrome The Budd-Chiari syndrome is a relatively rare disorder characterized by occlusion of the lumens of the hepatic veins with or without occlusion of the IVC lumen. The degree of occlusion and presence of collateral circulation predict the clinical course. Some patients die in the acute

Chapter 4  ■  The Liver   105

phase of liver failure. Causes of Budd-Chiari syndrome include coagulation abnormalities such as polycythemia rubra vera, chronic leukemia, and paroxysmal nocturnal hemoglobinuria; trauma; tumor extension from primary HCC, renal carcinoma, and adrenocortical carcinoma; pregnancy; congenital abnormalities; and obstructing membranes. The classic patient in North America is a young adult woman taking oral contraceptives who presents with an acute onset of ascites, right upper quadrant pain, hepatomegaly, and to a lesser extent, splenomegaly. In some cases, no etiologic factor is found. The syndrome is more common in other geographic areas, including India, South Africa, and Asia. Sonographic evaluation of the patient with BuddChiari syndrome includes gray-scale and Doppler features.104-115 Ascites is invariably seen. The liver is typically large and bulbous in the acute phase (Fig. 4-35, A). Hemorrhagic infarction may produce significant altered regional echogenicity. As infarcted areas become more fibrotic, echogenicity increases.105 The caudate lobe is often spared in Budd-Chiari syndrome because the emissary veins drain directly into the IVC at a lower level than the involved main hepatic veins. Increased blood flow through the caudate lobe leads to relative caudate enlargement. Real-time scanning allows the radiologist to evaluate the IVC and hepatic veins noninvasively. Sonographic features include evidence of the hepatic vein occlusion (Fig. 4-35, B, and Fig. 4-36) and the development of abnormal intrahepatic collaterals (Fig. 4-37). The extent of hepatic venous involvement in Budd-Chiari syndrome includes partial or complete inability to see the hepatic veins, stenosis with proximal dilation, intraluminal echogenicity, thickened walls, thrombosis (Figs.

A

4-38 and 4-39), and extensive intrahepatic collaterals107, 108 (see Fig. 4-37). Membranous “webs” may be identified as echogenic or focal obliterations of the lumen.108 Real-time ultrasonography, however, underestimates the presence of thrombosis and webs and may be inconclusive in a cirrhotic patient with hepatic veins that are difficult to image.107 Intrahepatic collaterals, on grayscale images, show as tubular vascular structures in an abnormal location and typically are seen extending from a hepatic vein to the liver surface, where they anastomose with systemic capsular vessels. Duplex Doppler ultrasound and color Doppler flow imaging (CDFI) can help determine both the presence and the direction of hepatic venous flow in the evaluation of patients with suspected Budd-Chiari syndrome. The middle and left hepatic veins are best scanned in the transverse plane at the level of the xiphoid process. From this angle, the veins are almost parallel to the Doppler beam, allowing optimal reception of their Doppler signals. The right hepatic vein is best evaluated from a right intercostal approach.110 The intricate pathways of blood flow out of the liver in the patient with Budd-Chiari syndrome can be mapped with documentation of hepatic venous occlusions, hepaticsystemic collaterals, hepatic venous–portal venous collaterals, and increased caliber of anomalous or accessory hepatic veins. The normal blood flow in the IVC and hepatic veins is phasic in response to both the cardiac and respiratory cycles.116 In Budd-Chiari syndrome, flow in the IVC, hepatic veins, or both, changes from phasic to absent, reversed, turbulent, or continuous.112,115 Continuous flow has been called the pseudoportal Doppler signal and appears to reflect either partial IVC obstruction or

B

FIGURE 4-35.  Acute Budd-Chiari syndrome. A, Transverse view of liver shows a large, bulbous caudate lobe. B, Sagittal view of right hepatic vein shows echoes within the vein lumen consistent with thrombosis, with absence of the vessel toward the inferior vena cava. Doppler ultrasound showed no flow in this vessel.

106   PART II  ■  Abdominal, Pelvic, and Thoracic Sonography

A

C

B

FIGURE 4-36.  Budd-Chiari syndrome. Abnormal hepatic vein appearance in three patients on transverse images of intrahepatic inferior vena cava. A, Right hepatic vein is not seen at all. Middle and left hepatic veins show tight strictures just proximal to the inferior vena cava. B, Right hepatic vein is seen as a thrombosed cord. Middle hepatic vein does not reach the inferior vena cava. Left hepatic vein is not seen. C, Only a single hepatic vein, the middle hepatic vein, can be seen as a thrombosed cord.

FIGURE 4-37.  Budd-Chiari syndrome. Abnormal intrahepatic collaterals on gray-scale sonograms in two patients. Both images show vessels with abnormal locations and increased tortuosity compared with the normal intrahepatic vasculature.

Chapter 4  ■  The Liver   107

A

C

B

D

FIGURE 4-38.  Budd-Chiari syndrome. A, Gray-scale transverse image of hepatic venous confluence shows complete absence of the right hepatic vein with obliteration of the lumen of a common trunk for the middle and left hepatic veins. B, Color Doppler image shows that blood flow in the middle hepatic vein (blue) is normally directed toward the inferior vena cava. As the trunk is obliterated, all the blood is flowing out of the left hepatic vein (red), which is abnormal. Other images showed anastomoses of the left hepatic vein with surface collaterals. C, Color Doppler image shows an anomalous left hepatic vein with flow to the inferior vena cava (normal direction) and aliasing from a long stricture. D, Spectral Doppler waveform of the anomalous left hepatic vein shows a very high abnormal velocity of approximately 140 cm/sec, confirming the tight stricture.

extrinsic IVC compression.111 The portal blood flow also may be affected and is characteristically either slowed or reversed.112 The addition of Doppler to gray-scale sonography in the patient with suspected Budd-Chiari syndrome lends strong supportive evidence to the gray-scale impression of missing, compressed, or otherwise abnormal hepatic veins and IVC.114,115 Associated reversal of flow in the portal vein and epigastric collaterals is also optimally assessed with this technique.115

Hepatic veno-occlusive disease causes progressive occlusion of the small hepatic venules. The disease is endemic in Jamaica, secondary to alkaloid toxicity from bush tea. In North America, most cases are iatrogenic, secondary to hepatic irradiation and chemotherapy used in bone marrow transplantation.113 Patients with hepatic veno-occlusive disease are clinically indistinguishable from those with Budd-Chiari syndrome. Duplex Doppler sonography demonstrates normal caliber, patency, and phasic forward (toward the heart)

108   PART II  ■  Abdominal, Pelvic, and Thoracic Sonography

IVC

A

B

RHV

C

D

FIGURE 4-39.  Budd-Chiari syndrome with extensive inferior vena cava thrombosis. A, Sagittal image of the inferior vena cava (IVC) shows that it is distended with echogenic thrombus. B, Middle hepatic vein as a thrombosed cord. C, Gray-scale image of right hepatic vein (RHV), and D, color Doppler image, show that anomalous right hepatic vein is distended with thrombus. There is flow in the vein proximal to the thrombus (blue).

flow of the main hepatic veins and IVC.113 Flow in the portal vein, however, may be abnormal, showing either reversed or “to and fro” flow.113,117 In addition, the diagnosis of hepatic veno-occlusive disease may be suggested in a patient with decreased portal blood flow (compared with baseline measurement before ablative therapy).113

proximally at the junction of the superior mesenteric and splenic veins and distally involving the portal venous radicles. The sonographic appearance is that of an anechoic cystic mass, which connects with the portal venous system. Pulsed Doppler sonographic examination demonstrates turbulent venous flow.118

Portal Vein Aneurysm

Intrahepatic Portosystemic Venous Shunts

Aneurysms of the portal vein are rare. Their origin is either congenital or acquired secondary to portal hypertension.118 Portal vein aneurysms have been described

Intrahepatic arterial-portal fistulas are well-recognized complications of large-gauge percutaneous liver biopsy and trauma. Conversely, intrahepatic portohepatic

Chapter 4  ■  The Liver   109

venous shunts are rare. Their cause is controversial and believed to be either congenital or related to portal hypertension.119,120 Patients typically are middle aged and present with hepatic encephalopathy. Anatomically, portohepatic venous shunts are more common in the right lobe. Sonography demonstrates a tortuous tubular vessel or complex vascular channels, which connect a branch of the portal vein to a hepatic vein or the IVC.118-121 The diagnosis is confirmed angiographically.

Hepatic Artery Aneurysm and Pseudoaneurysm The hepatic artery is the fourth most common site of an intra-abdominal aneurysm, following the infrarenal aorta, iliac, and splenic arteries. Eighty percent of patients with a hepatic artery aneurysm experience catastrophic rupture into the peritoneum, biliary tree, gastrointestinal tract, or portal vein.122 Hepatic artery pseudoaneurysm secondary to chronic pancreatitis has been described. The duplex Doppler sonographic examination revealed turbulent arterial flow within a sonolucent mass.122 Primary dissection of the hepatic artery is rare and in most cases leads to death before diagnosis.123 Sonography may show the intimal flap with the true and false channels.

Hereditary Hemorrhagic Telangiectasia Hereditary hemorrhagic telangiectasia, or Osler-WeberRendu disease, is an autosomal dominant disorder that causes arteriovenous (AV) malformations in the liver, hepatic fibrosis, and cirrhosis. Patients present with multiple telangiectasias and recurrent episodes of bleeding. Sonographic findings include a large feeding common hepatic artery up to 10 mm, multiple dilated tubular

A

structures representing AV malformations, and large draining hepatic veins secondary to AV shunting.124

Peliosis Hepatis Peliosis hepatis is a rare liver disorder characterized by blood-filled cavities ranging from less than a millimeter to many centimeters in diameter. It can be distinguished from hemangioma by the presence of portal tracts within the fibrous stroma of the blood spaces. The pathogenesis of peliosis hepatis involves rupture of the reticulin fibers that support the sinusoidal walls, secondary to cell injury or nonspecific hepatocellular necrosis.125 The diagnosis of peliosis can be made with certainty only by histologic examination. Most cases of peliosis affect the liver, although other solid internal organs and lymph nodes may be involved in the process as well. Although early reports described incidental detection of peliosis hepatis at autopsy in patients with chronic wasting disorders, it has now been seen following renal and liver transplantation, in association with a multitude of drugs, especially anabolic steroids, and with an increased incidence in HIV patients.126 The HIV association may occur alone or as part of bacillary angiomatosis in the spectrum of opportunistic infections of AIDS.127 Peliosis hepatis has the potential to be aggressive and lethal. The imaging features of peliosis hepatis have been described in single case reports,128-130 although often without adequate histologic confirmation. Angiographically, the peliotic lesions have been described as accumulations of contrast detected late in the arterial phase and becoming more distinct in the parenchymal phase.131 On sonography, described lesions are nonspecific and have shown single or multiple masses of heterogeneous echogenicity.128,129,132 Calcifications have been reported132 (Fig. 4-40). CT scans show low-attenuation nodular

B

FIGURE 4-40.  Peliosis hepatis. Peliosis hepatis in 34-year-old woman with deteriorating liver function necessitating transplantation. A, Sagittal right lobe, and B, sagittal left lobe, scans show multiple large liver masses with innumerable tiny punctate calcifications. (From Muradali D, Wilson SR, Wanless IR, et al. Peliosis hepatis with intrahepatic calcifications. J Ultrasound Med 1996;15:257-260.)

110   PART II  ■  Abdominal, Pelvic, and Thoracic Sonography

lesions that may or may not enhance with contrast injection.128,131 Peliosis hepatis is difficult to diagnose both clinically and radiologically and must be suspected in a susceptible individual with a liver mass.

HEPATIC MASSES Focal liver masses include a variety of malignant and benign neoplasms, as well as masses with developmental, inflammatory, and traumatic causes. In cross-sectional imaging, two basic issues relate to a focal liver lesion: characterization of a known liver lesion (what is it?) and detection (is it there?). The answer to either question requires a focused examination, often adjusted according to the clinical situation.

Liver Mass Characterization Characterization of a liver mass on conventional sonography is based on the appearance of the mass on grayscale imaging and vascular information derived from spectral, color, and power Doppler sonography. With excellent spatial and contrast resolution, the gray-scale morphology of a mass allows for the differentiation of cystic and solid masses, and characteristic appearances may suggest the correct diagnosis without further evaluation. More often, however, definitive diagnosis is not based on gray-scale information alone, but on vascular information obtained on conventional Doppler ultrasound examination. However, conventional Doppler often fails in the evaluation of a focal liver mass, particularly in a large patient or on a small or deep liver lesion, or on a mass with inherent weak Doppler signals. Motion artifact is also highly problematic for abdominal Doppler ultrasound studies, and a left lobe liver mass close to the pulsation of the cardiac apex, for example, is virtually always a failure for conventional Doppler. For these reasons, conventional ultrasound is not regarded highly for characterization of focal liver masses, and a mass detected on ultrasound is generally evaluated further with contrast-enhanced CT (CECT) or MRI for definitive characterization.

Role of Microbubble Contrast Agents Worldwide, noninvasive diagnosis of focal liver masses is achieved with CECT and MRI based on recognized enhancement patterns in the arterial and portal venous phases. These noninvasive methods of characterization have become so accurate that excisional and percutaneous biopsy for diagnosis of liver masses is now rarely performed. Over the last decade, however, contrastenhanced ultrasound has joined the ranks of CT and MRI in providing similar diagnostic information as well as information unique to CEUS.133

To address a failed Doppler ultrasound examination of a focal liver lesion, the two basic remedies are (1) inject a microbubble contrast agent to enhance the Doppler signal from blood and (2) use a specialized imaging technique such as pulse inversion sonography, which allows preferential detection of the signal from the contrast agent with suppression of the signal from background tissue. Ultrasound contrast agents currently in use are second-generation agents comprising tiny bubbles of a perfluorocarbon gas contained within a stabilizing shell. Microbubble contrast agents are blood pool agents that do not diffuse through the vascular endothelium. This is of potential importance when imaging the liver because comparable contrast agents for CT and MRI may diffuse into the interstitium of a tumor. Our personal experience with perfluorocarbon microbubble agents is largely based on the use of Definity (Lantheus Medical Imaging, Billerica, Mass) and brief exposure to Optison (GE Healthcare, Milwaukee).134,135 Microbubble contrast agents are approved for use in liver imaging in more than 70 countries. In our clinical practice, we routinely perform CEUS for characterization of incidentally detected liver masses, those found on surveillance scans of patients at risk for HCC, and any focal mass referred by our clinicians found on outside imaging or indeterminate on CT and MRI. In the United States, however, microbubble use in abdominal imaging has not yet been approved.136 Microbubble contrast agents for ultrasound are unique in that they interact with the imaging process.135 The major determinant of this interaction is the peak negative pressure of the transmitted ultrasound pulse, reflected by the mechanical index (MI), a number displayed on the ultrasound machine. The bubbles show stable, nonlinear oscillation when exposed to an ultrasound field with a low MI, with the production of harmonics of the transmitted frequency, including the frequency double that of the sound emitted by the transducer, the second harmonic. When the MI is raised sufficiently, the bubbles undergo irreversible disruption, with the production of a brief but bright, high-intensity ultrasound signal (see Chapter 3). Liver lesion characterization with microbubble contrast agents is based on lesional vascularity and lesional enhancement in the arterial and portal venous phase. Lesional vascularity assessment depends on continuous imaging of the agents while they are within the vascular pool. We document the presence, number, distribution, and morphology of any lesional vessels (Figs. 4-41 and 4-42). A low MI is essential because it will preserve the contrast agent population without destruction of the bubbles in the imaging field, allowing for prolonged periods of real-time observation. The morphology of the lesional vessels is discriminatory and facilitates the diagnosis of liver lesions.

Chapter 4  ■  The Liver   111

A

B

C

D

E

F

FIGURE 4-41.  Hepatocellular carcinoma. Characterization of a focal liver mass with microbubble contrast agents. A, Baseline gray-scale image shows a posterior focal mass that is hypoechoic. B, Taken at the same location with low mechanical index (MI), before arrival of microbubbles, the entire image now appears black. The lesion is not visible. C and D, Real-time images obtained with low MI. C, As the bubbles appear in the field of view, disorganized echogenic vessels are seen in the liver and in the lesion. D, Later in the arterial phase, more vessels are seen in the lesion than in the liver. E, Arterial phase image, at the peak of enhancement, shows the mass is hypervascular. F, Portal venous phase image shows that the liver is enhanced. The lesion is less echogenic than the liver or has “washed out.”

112   PART II  ■  Abdominal, Pelvic, and Thoracic Sonography

A

B

C

E

D

F

FIGURE 4-42.  Discriminatory features of vascular imaging with microbubble contrast agents. Left side, Baseline images; right side, vascular images. Top row, Hepatocellular carcinoma. A, Baseline shows an exoplytic mass in segment 6. B, Vessels in the anterior part of the lesion are tortuous and dysmorphic. Middle row, Focal nodular hyperplasia. C, Lesion is barely visible. D, Stellate vessels are classic for this diagnosis. Bottom row, Hemangioma. E, Baseline image shows the lesion is heterogeneous with a thin, echogenic border. F, Low-MI vascular image shows brightly enhanced peripheral nodules and pools. There are no visible linear vessels. (From Brannigan M, Burns PNB, Wilson SR. Blood flow patterns in focal liver lesions at microbubble enhanced ultrasound. Radiographics 2004; 24:921-935.)

Chapter 4  ■  The Liver   113

Lesional enhancement is best determined by comparing the echogenicity of the lesion to the echogenicity of the liver at a similar depth on the same frame and requires knowledge of liver blood flow. The liver has a dual blood supply from the hepatic artery and portal vein. The liver derives a larger proportion of its blood from the portal vein, whereas most liver tumors derive their blood supply from the hepatic artery. At the initiation of the injection, the low-MI technique will cause the entire field of view (FOV) to appear virtually black, regardless of the baseline appearance of the liver and the lesion in question. In fact, a known mass may be invisible at this point (see Fig. 4-41, B). As the microbubbles arrive in the FOV, the discrete vessels in the liver (Fig. 4-43) and then those within a liver lesion will be visualized, followed by increasing generalized enhancement as the microvascular volume of liver and lesion fills with the contrast agent. The liver parenchyma will appear more echogenic in the arterial phase than at baseline, and even more enhanced in the portal venous phase, as a reflection of its blood flow. Vascularity and enhancement patterns of a liver lesion, by comparison, will therefore reflect the actual blood flow and hemodynamics of the lesion in question, such that a hyperarterialized mass will appear more enhanced against a less enhanced liver

on an arterial phase sequence (see Fig. 4-41, E). Conversely, a hypoperfused lesion will appear as a dark or hypoechoic region within the enhanced liver on an arterial phase sequence. Currently, evaluation of lesional enhancement is usually performed with the low-MI technique just described. However, details of vessel morphology and lesional enhancement are even more sensitively assessed using a bubble-tracking technique called maximumintensity projection (MIP) imaging.137 In this technique, performed either at wash-in of contrast or at the peak of arterial phase enhancement, a brief high-MI exposure will destroy all the bubbles within the FOV. Sequential frames, as the lesion and liver are reperfused, track the bubble course by adding information between sequential frames. There are established algorithms for the diagnosis of focal liver masses with CEUS, with similarities to CT and MR algorithms but also important differences138-140 (Table 4-3). Diagnosis of benign liver masses, hemangioma, and focal nodular hyperplasia (FNH) is close to 100%, showing characteristic features of enhancement in the arterial phase and sustained enhancement in the portal venous phase, such that their enhancement equals or exceeds the enhancement of the adjacent liver. Malignant tumors, by comparison, tend to show washout, such that the tumor appears unenhanced in the portal venous phase of enhancement (see Fig. 4-41, F). Exceptions to this general rule include frequent washout of benign hepatic adenoma and delayed or no washout of HCC. Discrimination of benign and malignant liver masses has similarly high accuracy.141

Liver Mass Detection

FIGURE 4-43.  Normal liver vasculature. Temporal maximum-intensity projection image shows accumulated enhancement in 11 seconds after contrast material arrives in liver. Unprecedented depiction of vessel structure to fifth-order branching is evident. Focal unenhanced region (arrow) is slowly perfusing hemangioma. (From Wilson SR, Jang HJ, Kim TK, et al. Real-time temporal maximum-intensity-projection imaging of hepatic lesions with contrast-enhanced sonography. AJR Am J Roentgenol 2008;190:691-695.)

Contrary to popular belief, excellent spatial resolution allows small lesions to be well seen on sonography. Therefore it is not size but echogenicity that determines lesion conspicuity on a sonogram. That is, a tiny mass of only a few millimeters will be easily seen if it is increased or decreased in echogenicity compared with the adjacent liver parenchyma. Because many metastases are either hypoechoic or hyperechoic relative to the liver, a careful examination should allow for their detection. Nonetheless, many metastatic lesions are of similar echogenicity to the background liver, making their detection difficult or impossible, even if they are of a substantial size. This occurs when the backscatter from the lesion is virtually identical to the backscatter from the liver parenchyma. To combat this inherent problem of lack of contrast between many metastatic liver lesions and the background liver on conventional sonography, the most effective method to date to improve lesion visibility is to perform contrast-enhanced liver ultrasound (Fig. 4-44). The two methods available both produce enhancement of the background liver without enhancement of the

114   PART II  ■  Abdominal, Pelvic, and Thoracic Sonography

TABLE 4-3.  SCHEMATIC OF ALGORITHM FOR LIVER MASS DIAGNOSIS ON CEUS

Hemangioma

or

AP Peripheral nodular enhancement Centripetal progression of enhancement PVP Complete or partial fill-in AP Centrifugal hypervascular enhancement Stellate arteries PVP Sustained enhancement Hypoechoic central scar

FNH

AP Diffuse or centripetal hypervascular enhancement Dysmorphic arteries

Adenoma

or

AP Rim enhancement Diffuse hypervascular Hypovascular

or

Metastases

or

PVP Fast washout

Arterial phase (AP) (+) Enhancement

PVP Sustained enhancement Soft wash out

Portal venous phase (PVP) Soft wash out

(–) enhancement (wash out)

From Wilson SR, Burns PN. Microbubble contrast enhanced ultrasound in body imaging: what role? Radiology 2010. FNH, Focal nodular hyperplasia.

metastatic lesions, thereby improving their conspicuity. Although their mechanism of action is different, in both there is microbubble enhancement of the normal liver with no enhancement of the liver metastases. This increases the backscatter from the liver compared with the liver lesions, thereby improving their detection. The first method used the first-generation contrast agent Levovist (Schering AG, Berlin). After clearance of the contrast agent from the vascular pool, the microbubble persisted in the liver, probably within the Kupffer cells on the basis of phagocytosis. A high-MI sweep through the liver produced bright enhancement in the distribution of the bubbles. Therefore, all normal liver enhances. Liver metastases, lacking Kupffer cells, do not enhance and therefore show as black or hypoechoic holes within the enhanced parenchyma.142 In a multicenter study conducted in Europe and Canada in which we participated, more and smaller lesions were seen than on baseline scan.143 Overall, lesion detection was equivalent

to CT and MRI. The decibel difference between the lesions and the liver parenchyma is increased many fold because of increased backscatter from contrast agent within the normal liver tissue. Although many results were compelling, these first-generation contrast agents are no longer marketed. Therefore, current requirements for improved lesion detection use a second technique of CEUS with a perfluorocarbon contrast agent and low-MI scanning in both the arterial and the portal venous phase. The use of a low-MI imaging technique for lesion detection has advantages in terms of scanning because the microbubble population is preserved and timing is not so critical. Virtually all metastases and HCCs will be unenhanced relative to the liver in the portal phase because the liver parenchyma is optimally enhanced in this phase. Therefore, all malignant lesions tend to appear hypoechoic in the portal phase, allowing for improved lesion detection (see Figs. 4-41, F, and 4-44, C and D). This observation,

Chapter 4  ■  The Liver   115

A

B

C

D

FIGURE 4-44.  Improved detection of focal liver masses with microbubble contrast agents. A and B, Levovist (Schering, Berlin). A, Baseline sonogram shows a subtle isoechoic mass with a hypoechoic halo. B, Postvascular image shows increased echogenicity in the liver. The lesion is strikingly hypoechoic and has increased conspicuity. C and D, Definity (Lantheus Medical Imaging, Billerica, Mass). C, Baseline sonogram does not show any metastatic lesions in this patient with lung carcinoma. D, Portal venous phase image shows multiple focal unenhanced metastases.

that malignant lesions tend to be hypoechoic in the portal venous phase of perfluorocarbon liver enhancement, is helpful for both lesion detection and lesion characterization. Enhancement of benign lesions, FNH, and hemangioma generally equals or exceeds liver enhancement in the portal venous phase. Detection of hypervascular liver masses (e.g., HCC, metastases) is also improved by scanning with perfluorocarbon agents in the arterial phase. These agents will show as hyperechoic masses relative to the liver parenchyma in the arterial phase because they are predominantly supplied by hepatic arterial flow.

HEPATIC NEOPLASMS Sonographic visualization of a focal liver mass may occur in a variety of clinical scenarios, ranging from incidental

detection to identification in a symptomatic patient or as part of a focused search in a patient at risk for hepatic neoplasm. Hemangiomas, FNH, and adenomas are the benign neoplasms typically encountered in the liver, whereas HCC and metastases account for the majority of malignant tumors. The role of medical imaging in the evaluation of an identified focal liver mass is to determine which masses are significant, requiring confirmations of their diagnoses, and which masses are likely to be insignificant and benign, not requiring further evaluation to confirm their nature. On a sonographic study, there is considerable overlap in the appearances of focal liver masses. Once a liver mass is seen, however, the excellent contrast and spatial resolution of state-of-the-art ultrasound equipment have provided guidelines for the initial management of patients,144 which include recognition of the following features:

116   PART II  ■  Abdominal, Pelvic, and Thoracic Sonography

• A hypoechoic halo identified around an echogenic or isoechoic liver mass is an ominous sonographic sign necessitating definitive diagnosis. • A hypoechoic and solid liver mass is highly likely to be significant and also requires definitive diagnosis. • Multiple solid liver masses may be significant and suggest metastatic or multifocal malignant liver disease. However, hemangiomas are also frequently multiple. • Clinical history of malignancy, chronic liver disease or hepatitis, and symptoms referable to the liver are requisite information for interpretation of a focal liver lesion.

Benign Hepatic Neoplasms Cavernous Hemangioma Cavernous hemangiomas are the most common benign tumors of the liver, occurring in approximately 4% of the population. They occur in all age groups but are more common in adults, particularly women, with a female/male ratio of approximately 5 : 1.145 The vast majority of hemangiomas are small, asymptomatic, and discovered incidentally. Large lesions may rarely produce symptoms of acute abdominal pain, caused by hemorrhage or thrombosis within the tumor. Thrombocytopenia, caused by sequestration and destruction of platelets within a large cavernous hemangioma (Kasabach-Merritt syndrome), occasionally occurs in infants and is rare in adults. Traditional teaching suggests that once identified in the adult, hemangiomas usually have reached a stable size, rarely changing in appearance or size.146,147 In our practice, however, we have documented substantial growth of some lesions over many years of follow-up. Hemangiomas may enlarge during pregnancy or with the administration of estrogens, suggesting the tumor is hormone dependent. Histologically, hemangiomas consist of multiple vascular channels that are lined by a single layer of endothelium and separated and supported by fibrous septa. The vascular spaces may contain thrombi. The sonographic appearance of cavernous hemangioma varies. Typically the lesion is small (50% involved, all cores less than Gleason 7, fewer than three positive cores) and PSA density less than 0.15.119 Watchful waiting is used in men who have asymptomatic cancer but are unlikely to benefit from therapy because of comorbid conditions. They are monitored until they become symptomatic and then receive palliative care, usually with hormones.50 Active surveillance (active monitoring with curative intent) is an increasingly popular “PSA era phenomenon” that is almost unique to prostate cancer. It recognizes that many men with low-risk cancer will never suffer from the cancer and will die from other causes. After initial diagnosis of low-risk disease, they are actively monitored with PSA, DRE, TRUS, and repeat biopsy to detect signs of progression before undergoing therapy. Strict criteria are used to define suitable low-risk patients (PSA 4) may be present and result from the separation of parathyroid anlage when the glands pull away from the pouch structures during the embryologic branchial complex phase.9,10 These supernu-

merary glands are often associated with the thymus in the anterior mediastinum, suggesting a relationship in their development with the inferior parathyroid glands.11 Supernumerary glands have been reported in 13% of the population at autopsy studies;3,4 however, many of these are small, rudimentary or split glands. “Proper” supernumerary glands (>5 mg and located well away from the other four glands) are found in 5% of cases. The presence of fewer than four parathyroid glands is rare clinically, but has been reported in 3% at autopsy. Normal parathyroid glands vary from a yellow to a red-brown color, depending on the degree of vascularity and the relative content of yellow parenchymal fat and chief cells.8 The chief cells are the primary source for the production of parathyroid hormone (PTH, parathormone). The percentage of glandular fat typically increases with age or with disuse atrophy. Hyperfunctioning glands resulting from adenomas or hyperplasia contain relatively little fat and are vascular, thus more reddish. The glands are generally oval or bean shaped but may be spherical, lobular, elongated, or flattened. Although normal parathyroid glands are occasionally seen with high-frequency ultrasound,12,13 typically they are not visualized, likely because of their small size, deep location, and poor conspicuity related to increased glandular fat. Eutopic parathyroid glands typically derive their major blood supply from branches of the inferior thyroid artery, with a lesser and variable contribution to the superior glands from the superior thyroid artery.3,7

PRIMARY HYPERPARATHYROIDISM Prevalence Primary hyperparathyroidism is now recognized as a common endocrine disease, with prevalence in the United States of 1 to 2 per 1000 population.14 Women are affected two to three times more frequently than men, particularly after menopause. More than half of patients with primary hyperparathyroidism are over 50 years old, and cases are rare in those under age 20.

Diagnosis Primary hyperparathyroidism is usually suspected because an increased serum calcium level is detected on routine biochemical screening. Elevated ionized serum calcium level, hypophosphatasia, and hypercalciuria may be further biochemical clues to the disease. A serum PTH level that is “inappropriately high” for the corresponding serum calcium level confirms the diagnosis. Even when the PTH level is within the upper limits of the normal range in a hypercalcemic patient, the diagnosis of primary hyperparathyroidism should still be suspected, since hypercalcemia from other nonparathyroid

752   PART III  ■  Small Parts, Carotid Artery, and Peripheral Vessel Sonography

causes (including malignancy) should suppress the glandular function and decrease the serum PTH level. Because of earlier detection by increasingly routine laboratory tests, the later “classic” signs of hyperparathyroidism, such as “painful bones, renal stones, abdominal groans, and psychic moans,” are often not present. Many patients are now diagnosed before severe manifestations of hyperparathyroidism, such as nephrolithiasis, osteopenia, subperiosteal resorption, and osteitis fibrosis cystica. In general, patients rarely have obvious symptoms unless their serum calcium level exceeds 12 mg/dL. However, subtle nonspecific symptoms, such as muscle weakness, malaise, constipation, dyspepsia, polydipsia, and polyuria, may be elicited from these otherwise asymp­tomatic patients by more specific questioning.

Pathology Primary hyperparathyroidism is caused by a single adenoma in 80% to 90% of cases, by multiple gland enlargement in 10% to 20%, and by carcinoma in less than 1%.6,15,16 A solitary adenoma may involve any one of the four glands. Multigland enlargement most often results from primary parathyroid hyperplasia and less often from multiple adenomas. Hyperplasia usually involves all four glands asymmetrically, whereas multiple adenomas may involve two or possibly three glands. An adenoma and hyperplasia cannot always be reliably distinguished histologically, and the sample may be referred to as “hypercellular parathyroid” tissue. Because of this inconsistent pattern of gland involvement, and because distinguishing hyperplasia from multiple adenomas is difficult pathologically, these two entities are often histologically considered together as “multiple gland disease.”17 Most cases of primary hyperparathyroidism are sporadic. However, prior external neck irradiation has been associated with the development of hyperparathyroidism in a small percentage of cases. Patients receiving long-term lithium therapy may also present with primary hyperparathyroidism. Up to 10% of cases may occur on a hereditary basis, most often caused by multiple endocrine neoplasia syndrome, type I (MEN I). This condition is an uncommon disorder that most typically follows an autosomal dominant pattern of inheritance and has a high penetrance, resulting in adenomatous parathyroid hyperplasia, as well as pancreatic islet cell tumors and pituitary adenomas. Multiple–parathyroid

CAUSES OF PRIMARY HYPERPARATHYROIDISM Single adenoma Multiple gland disease Carcinoma

80%-90% 10%-20% 14 mg/dL). The diagnosis is often made at operation when the surgeon discovers an enlarged, firm gland that is adherent to the surrounding tissues due to local invasion. A thick, fibrotic capsule is often present. Treatment consists of en bloc resection without entering the capsule, to prevent tumor seeding. In many cases, cure may not be possible because of the invasive and metastatic nature of the disease. Generally, death occurs not from tumor spread, but from complications associated with unrelenting hyperparathyroidism.

Treatment No effective definitive medical therapies are available for the treatment of primary hyperparathyroidism. Short-term hypocalcemic agents include calcitonin and the bisphosphonates. Calcimimetics (calcium-sensing receptor agonists) such as cinacalcet and synthetic vitamin D analogs such as paricalcitol are used mainly in the treatment of secondary hyperparathyroidism. Surgery is the only definitive treatment for primary hyperparathyroidism. Studies demonstrate that surgical cure rates by an experienced surgeon are greater than 95%, and the morbidity and mortality rates are extremely low.24,25 Therefore, in symptomatic patients with primary hyperparathyroidism, the treatment of choice is surgical excision of the involved parathyroid gland or glands. However, now that many cases of primary hyper­ parathyroidism are discovered in the early stages of the disease, some controversy exists as to whether asymptomatic patients with minimal hypercalcemia should be treated surgically, or followed medically with frequent

Chapter 19  ■  The Parathyroid Glands   753

measurements of bone density, serum calcium levels, and urinary calcium excretion and monitoring for nephrolithiasis. A prospective 10-year clinical follow-up study reported that of 52 asymptomatic patients with primary hyperparathyroidism with calcium levels less than 11 mg/ dL, 73% did well, with no evidence of disease progression. However, 27% had evidence of progression based on development of one or more indications for surgery.26 Recommendations for the management of asymptomatic primary hyperparathyroidism have been outlined in various articles, many of which are based on the National Institutes of Health (NIH) Consensus Conference statement and its subsequent updates. However, this area continues to evolve, and approaches to treatment may differ among clinical practices.24,25,27-30

tioning parathyroid lipoadenomas are more echogenic than the adjacent thyroid gland because of their high fat content34 (Fig. 19-3, G). A great majority of parathyroid adenomas are homogeneously solid. About 2% have internal cystic components resulting from cystic degeneration (most often) or true simple cysts (less often).35,36 Adenomas may rarely contain internal calcification (Fig. 19-3, H and I).

Vascularity Color flow, spectral, and power Doppler sonography of an enlarged parathyroid gland may demonstrate a hypervascular pattern with prominent diastolic flow (Fig. 19-4). An enlarged extrathyroidal artery, often originating from branches of the inferior thyroidal artery, may be visualized supplying the adenoma with its insertion along the long-axis pole.37-42 A finding described in parathyroid adenomas is a vascular arc, which envelops 90 to 270 degrees of the mass. This vascular flow pattern may increase the sensitivity of initial detection of parathyroid adenomas and aid in confirming the diagnosis by allowing for differentiation from lymph nodes, which have a central hilar flow pattern. Asymmetric increased vascular flow may also be present in the thyroid gland adjacent to a parathyroid adenoma.

SONOGRAPHIC APPEARANCE Shape Parathyroid adenomas are typically oval or bean shaped (Fig. 19-2). As parathyroid glands enlarge, they dissect between longitudinally oriented tissue planes in the neck and acquire a characteristic oblong shape. If this process is exaggerated, they can become tubular or flattened. There is often asymmetry in the enlargement, and the cephalic and/or caudal end can be more bulbous, producing a triangular, tapering, teardrop or bilobed shape.19,31-33

Size Most parathyroid adenomas are 0.8 to 1.5 cm long and weigh 500 to 1000 mg. The smallest adenomas can be minimally enlarged glands that appear virtually normal during surgery but are found to be hypercellular on pathologic examination (Fig. 19-5; Video 19-1). Large adenomas can be 5 cm or more in length and weigh more than 10 g. Preoperative serum calcium levels are usually higher in patients with larger adenomas.31

Echogenicity and Internal Architecture The echogenicity of most parathyroid adenomas is substantially less than that of normal thyroid tissue (Fig. 19-3). The characteristic hypoechoic appearance of parathyroid adenomas is caused by the uniform hypercellularity of the gland with little fat content, which leaves few interfaces for reflecting sound. Occasionally, adenomas have a heterogeneous appearance, with areas of increased and decreased echogenicity. The rare, func-

T J

A

Tr

Multiple Gland Disease Multiple gland disease may be caused by diffuse hyperplasia or multiple adenomas. Individually, these enlarged

T

T

T

C C

B

FIGURE 19-2.  Typical parathyroid adenoma. A, Transverse, and, B, longitudinal, sonograms of a typical adenoma (arrows) located adjacent to the posterior aspect of the thyroid (T); Tr, trachea; C, common carotid artery; J, internal jugular vein.

754   PART III  ■  Small Parts, Carotid Artery, and Peripheral Vessel Sonography

A

B

C

T T J

D

E

F

T

T

T

G

H

I

FIGURE 19-3.  Spectrum of echogenicity and internal architecture of parathyroid adenomas and enlarged hyperplastic glands. Longitudinal sonograms. A, Typical homogeneous hypoechoic appearance of a parathyroid adenoma (arrows) with respect to the overlying thyroid tissue. B, Highly hypoechoic solid adenoma (arrow). C, Mixed-geographic echogenicity. The adenoma (arrows) is hyperechoic in its cranial portion and hypoechoic in its caudal portion. D, An adenoma (arrows) with diffusely heterogeneous echotexture adenoma (arrows); T, thyroid. E, Partial cystic change. An ectopic adenoma (arrows) posterior to the jugular vein (J) has both solid and cystic components. F, Completely cystic 2-cm adenoma (cursors) near the lower pole of the thyroid (T). G, A lipoadenoma (arrow) is more echogenic than the adjacent lower pole thyroid tissue (T). H, Enlarged parathyroid gland (arrows) with small, nonshadowing calcifications in the setting of secondary hyperparathyroidism related to chronic renal failure; T, thyroid. I, Enlarged parathyroid gland (arrows) with densely shadowing peripheral calcifications in the setting of secondary hyperparathyroidism; T, thyroid.

glands may have the same sonographic and gross appearance as other parathyroid adenomas (Fig. 19-6; Video 19-2). However, the glands may be inconsistently and asymmetrically enlarged, and the diagnosis of multigland disease can be difficult to make sonographically. For

example, if one gland is much larger than the others, the appearance may be misinterpreted as solitary adenomatous disease. Alternatively, if multiple glands are only minimally enlarged, the diagnosis may be missed altogether.

Chapter 19  ■  The Parathyroid Glands   755

A

B

FIGURE 19-4.  Typical hypervascularity of parathyroid adenoma. A, Longitudinal gray-scale, and B, longitudinal power Doppler ultrasound images show hypervascularity of a parathyroid adenoma with polar feeding vessel and prominent peripheral vascular arcs. C, Longitudinal power Doppler sonogram in another patient shows larger parathyroid adenoma with polar feeding vessel and prominent peripheral vascularity.

C

A

B

C

FIGURE 19-5.  Spectrum of size of parathyroid adenomas. Longitudinal sonograms. A, Minimally enlarged, 0.5 × 0.2–cm parathyroid adenoma (cursors). B, Typical midsized, 1.5 × 0.6–cm, 400-mg adenoma (arrow). C, Large, 3.5 × 2–cm, >4000-mg adenoma (cursors).

756   PART III  ■  Small Parts, Carotid Artery, and Peripheral Vessel Sonography

Tr

T C

C

A

B

FIGURE 19-6.  Multiple gland disease. A, Longitudinal sonogram of the right neck shows superior and inferior parathyroid gland enlargement (arrows) in the setting of secondary hyperparathyroidism, which can be difficult to distinguish from multiple adenomas; T, thyroid. B, Transverse sonogram in another patient shows enlargement of bilateral superior parathyroid glands (arrows) in the setting of secondary hyperparathyroidism; Tr, trachea; C, common carotid artery.

Carcinoma Carcinomas are usually larger than adenomas.43-45 Carcinomas often measure more than 2 cm, versus about 1 cm for adenomas (Fig. 19-7). On ultrasound, carci­ nomas also frequently have a lobular contour, hetero­ geneous internal architecture, and internal cystic components. However, large adenomas may also have these features. In many cases, prospective carcinomas are indistinguishable sonographically from large, benign adenomas.43 Some authors report that a depth/width ratio of 1 or greater is a sonographic feature more associated with carcinoma rather than adenoma, with sensitivity and specificity of 94% and 95%, respectively.45 Gross evidence of invasion of adjacent structures, such as vessels or muscles, is a reliable preoperative sonographic criterion for the diagnosis of malignancy, but this is an uncommon finding.

ADENOMA LOCALIZATION Sonographic Examination and Typical Locations The sonographic examination of the neck for parathyroid adenoma localization is performed with the patient supine. The patient’s neck is hyperextended by a pad centered under the scapulae, and the examiner usually sits at the patient’s head. High-frequency transducers (8-17 MHz) are used to provide optimal spatial resolution and visualization in most patients; the highest frequency possible should be used that still allows for tissue penetration to visualize the deeper structures, such as the longus colli muscles. In obese patients with thick necks or with large multinodular thyroid glands, use of a 5-MHz to 8-MHz transducer may be necessary to obtain adequate depth of penetration. The pattern of the sonographic survey of the neck for adenoma localization can be considered in terms of the

pattern of dissection and visualization that the surgeon uses in a thorough neck exploration. The typical superior parathyroid adenoma is usually adjacent to the posterior aspect of the midportion of the thyroid (Fig. 19-8; Video 19-3). The location of the typical inferior parathyroid adenoma is more variable but usually lies close to the lower pole of the thyroid (Fig. 19-9; Video 19-4). Most of these inferior adenomas are adjacent to the posterior aspect of the lower pole of the thyroid, and the rest are in the soft tissues 1 to 2 cm inferior to the thyroid. Therefore the examination is initiated on one side of the neck, centered in the region of the thyroid gland, with the electronic focus placed deep to the thyroid. High-resolution gray-scale images are obtained in the transverse (axial) and longitudinal (sagittal) planes. Any potential parathyroid adenomas detected in the transverse scan plane must be confirmed by longitudinal imaging to prevent mistaking other structures for an adenoma. Some authors recommend the use of compression of the superficial soft tissues to aid in adenoma detection.12,41 This has been described as “graded” compression with the transducer to effect minimal deformity of the overlying subcutaneous tissues and strap muscles and increase the conspicuity of deeper, smaller adenomas (2% to 1.5 cm) with prominent color flow that present without clinical symptoms of inflammation are more likely to be malignant. Metastatic tumors to the epididymis are also rare. The most common primary sites include the testicle, stomach, kidney, prostate, colon, and less often the pancreas122,123 (Fig. 21-24, I).

Epididymal Lesions Sperm Granuloma Sperm granulomas are thought to arise from extravasation of spermatozoa into the soft tissues surrounding the epididymis, producing a necrotizing granulomatous response.1 These lesions may be painful or asymptomatic, and they are most often found in patients after vasectomy. Sperm granulomas may also be associated with prior epididymal infection or trauma. The typical sonographic appearance is that of a solid, hypoechoic or heterogeneous mass, usually located in the epididymis, although it may simulate an intratesticular lesion (Fig. 21-24, B). Chronic sperm granuloma may contain calcification.124

Fibrous Pseudotumor Fibrous pseudotumor is a rare, nonneoplastic mass of reactive fibrous tissue that may involve the tunica vaginalis or epididymis. On sonography, fibrous pseudotumors may appear as hypoechoic, hyperechoic, or heterogeneous paratesticular masses125-127 (Fig. 21-24, C ).

Cystic Lesions Spermatoceles are more common than epididymal cysts. Both were seen in 20% to 40% of all asymptomatic patients studied by Leung et al.,128 and 30% were multiple cysts. Both epididymal cysts and spermatoceles are thought to result from dilation of the epididymal tubules, but the contents of these masses differ.10 Cysts contain clear serous fluid, whereas spermatoceles are filled with spermatozoa and sediment containing lymphocytes, fat globules, and cellular debris, giving the fluid a thick, milky appearance.1 Both lesions may result from prior episodes of epididymitis or trauma. Spermatoceles and epididymal cysts appear identical on sonography: anechoic, circumscribed masses with no or few internal echoes; loculations and septations are often seen (Fig. 21-25). Rarely, a spermatocele may be hyper-

echoic.5 Differentiation between a spermatocele and an epididymal cyst is rarely important clinically. Spermatoceles almost always occur in the head of the epididymis, whereas epididymal cysts arise throughout the length of the epididymis.

Postvasectomy Changes in Epididymis Sonographic changes in the epididymis are very common in patients after vasectomy.129,130 These findings include epididymal enlargement with tubular ectasia and the development of sperm granulomas and cysts (Fig. 21-26; Video 21-4). It is assumed that vasectomy produces increased pressure in the epididymal tubules, causing tubular rupture with subsequent formation of sperm granulomas. The dilated vas deferens may be seen in addition to the dilated epididymis. An unusual appearance described as “dancing megasperm” is occasionally seen in patients with vasectomy (Video 21-5). High reflective echoes within the dilated epididymis appear to move independently, shown histologically to be aggregations of spermatozoa and macrophages.131

Chronic Epididymitis Patients with incompletely treated acute bacterial epididymitis usually present with a chronically painful scrotal mass (Fig. 21-24, A). Patients with chronic granulomatous epididymitis caused by spread of tuberculosis from the genitourinary tract complain of a hard, nontender scrotal mass.10 Sonography most often shows a thickened tunica albuginea and a thickened, irregular epididymis (Fig. 21-27). Calcification may be identified within the tunica albuginea or epididymis.1 Untreated granulomatous epididymitis will spread to the testes in 60% to 80% of cases. Focal testicular involvement may simulate the appearance of a testicular neoplasm on sonography.

ACUTE SCROTAL PAIN The differential diagnosis of an acutely painful and swollen scrotum includes torsion of the spermatic cord and testis, torsion of a testicular appendage, epididymitis or orchitis, acute hydrocele, strangulated hernia, idiopathic scrotal edema, Henoch-Schönlein purpura, abscess, traumatic hemorrhage, hemorrhage into a testicular neoplasm, and scrotal fat necrosis. Torsion of the spermatic cord and acute epididymitis or epidi­ dymo-orchitis are the most common causes of acute scrotal pain. These entities cannot be distinguished by physical examination or laboratory tests in up to 50% of patients.132 Immediate surgical exploration has been advised in boys and young men with acute scrotal pain, unless a definitive diagnosis of epididymitis or orchitis can be made. This aggressive approach has resulted in an

Chapter 21  ■  The Scrotum   865

A

B

C

D

E

F

FIGURE 21-25.  Extratesticular scrotal cysts: spectrum of appearances. A, Spermatocele. Longitudinal scan shows an anechoic cyst in head of the epididymis. B, Spermatocele. Longitudinal scan shows a large cyst containing internal echoes in head of the epididymis. C, Septate spermatocele. Longitudinal scan shows a septate cyst in head of the epididymis. D, Epididymal cyst. Longitudinal scan shows a cyst in body of the epididymis. E, Cyst of vas deferens remnant. Longitudinal scan shows a cyst with internal echoes inferior to the testis (surgically proven). F, Epidermoid inclusion cyst of epididymis. Longitudinal color Doppler scan shows bilobed cystic mass in head of the epididymis surrounded by vessels.

866   PART III  ■  Small Parts, Carotid Artery, and Peripheral Vessel Sonography

CAUSES OF ACUTE SCROTAL PAIN Torsion of the testis Epididymo-orchitis Testicular appendage torsion Strangulated hernia Idiopathic scrotal edema Trauma Henoch-Schönlein purpura

increased testicular salvage rate from torsion, but also an increase in unnecessary surgical procedures. Testicular radionuclide scintigraphy, MRI, real-time sonography, and Doppler sonography have been used to increase the accuracy of distinguishing between infection and torsion.133 Currently, sonography using color flow or

FIGURE 21-26.  Postvasectomy change in epididymis. Longitudinal image of the scrotum shows tubular ectasia of the epididymis in a patient who had a vasectomy.

A

power Doppler is the imaging study of choice to diagnose the cause of acute scrotal pain.

Torsion Torsion is more common in boys than in men, and it represents only 20% of the acute scrotal pathologic phenomena in postpubertal males.1 However, prompt diagnosis is necessary because torsion requires immediate surgery to preserve the testis. The testicular salvage rate is 80% to 100% if surgery is performed within 5 to 6 hours of the onset of pain, 70% if surgery is performed within 6 to 12 hours, and only 20% if surgery is delayed for more than 12 hours.134 There are two types of testicular torsion: intravaginal and extravaginal. Intravaginal torsion is the more common type, occurring most frequently at puberty. It results from anomalous suspension of the testis by a long stalk of spermatic cord, resulting in complete investment of the testis and epididymis by the tunica vaginalis. This anomaly has been likened to a bell-clapper (Fig. 21-28). Anomalous testicular suspension is bilateral in 50% to 80% of patients. There is a tenfold greater incidence of torsion in undescended testes after orchiopexy. Extravaginal torsion most often occurs in newborns without the “bell clapper” deformity. It is thought to result from a poor or absent attachment of the testis to the scrotal wall, allowing rotation of the testis, epididymis, and tunica vaginalis as a unit and causing torsion of the cord at the level of the external ring135,136 (Fig. 21-28, D). The more compliant veins are obstructed before the arteries in both forms of torsion, resulting in early vascular engorgement and edema of the testicle.

B

FIGURE 21-27.  Tuberculous epididymo-orchitis. A, Longitudinal scan shows a heterogeneous mass with calcification involving the head and body of the epididymis and the adjacent testis (T). B, Longitudinal color Doppler image shows increased vascularity in the epididymis and adjacent testis.

Chapter 21  ■  The Scrotum   867

A

B

C

D

FIGURE 21-28.  “Bell clapper” anomaly, intravaginal torsion, and extravaginal torsion. A, Normal anatomy. The tunica vaginalis (arrows) does not completely surround the testis and epididymis, which are attached to the posterior scrotal wall (short arrow). B, Bell-clapper anomaly. The tunica vaginalis (arrows) completely surrounds the testis, epididymis, and part of the spermatic cord, predisposing to torsion. C, Intravaginal torsion. Bell-clapper anomaly with complete torsion of the spermatic cord, compromising the blood supply to the testis. D, Extravaginal torsion in a neonate. Tunica vaginalis (arrows) is in normal position, but abnormal motility allows rotation of the testis, epididymis, and spermatic cord.

Several gray-scale sonographic changes occur in the acute phase of torsion, within 1 to 6 hours.132,137,138 Initially the testis becomes enlarged, with a normal echogenicity, and later it becomes heterogeneous and hypoechoic compared with the contralateral normal testis139-141 (Fig. 21-29). A hypoechoic or heterogeneous echogenicity may indicate nonviability.142 Generalized testicular hyperechogenicity has been reported in the absence of histologic changes of testicular hemorrhage or infarction.140 Torsion may change the position of the long axis of the testis (Fig. 21-29, B). Extratesticular sonographic findings typically occur in torsion and are important to recognize. The spermatic cord immediately cranial to the testis and epididymis is twisted, causing a characteristic torsion knot or “whirlpool pattern” of concentric layers

seen on sonography or MRI137,143,144 (Fig. 21-29, G and H). The epididymis may be enlarged and heterogeneous because of hemorrhage and may be difficult to separate from the torsion knot of the spermatic cord. This spherical epididymis-cord complex can be mistaken for epididymitis.137 A reactive hydrocele and scrotal skin thickening are often seen with torsion. Large, echogenic or complex extratesticular masses caused by hemorrhage in the tunica vaginalis or epididymis may be seen in patients with undiagnosed torsion.145 The gray-scale findings of acute and subacute torsion are not specific and may be seen in testicular infarction caused by epididymitis, epididymo-orchitis, and traumatic testicular rupture or infarction. Color Doppler sonography is the most useful and most rapid technique to establish the diagnosis of testicular torsion and to help distinguish torsion from epididymo-orchitis132,139,146 (Fig. 21-29). In torsion, blood flow is absent in the affected testicle or significantly less than in the normal, contralateral testicle. Meticulous scanning of the testicular parenchyma with the use of low-flow detection Doppler settings (low pulse repetition frequency, low wall filter, high Doppler gain) is important because testicular vessels are small and have low flow velocities, especially in prepubertal boys. Color flow Doppler sonography is more sensitive for showing decreased testicular flow in incomplete torsion than is nuclear scintigraphy.147 Power Doppler and frequency shift color Doppler sonography are used, although the techniques appear to have equivalent sensitivity in the diagnosis of torsion.148-153 In testicular torsion, color Doppler sonography has a sensitivity of 80% to 98%, a specificity of 97% to 100%, and an accuracy rate of 97%.137,146,154 The use of intravascular contrast agents in sonography may improve the sensitivity of detecting blood flow in the scrotum, but this has not yet been proved in a large series.150 In pediatric patients, it may be difficult to document flow in a normal testis.155 In practice, many surgeons elect to explore the testis surgically if clinical symptoms and signs are suggestive and results of the sonographic examination are equivocal. Potential pitfalls in using sonography in the diagnosis of torsion are partial torsion, torsion/detorsion, and ischemia from orchitis. Torsion of at least 540 degrees is necessary for complete arterial occlusion.146,156 With partial torsion of 360 degrees, or less, arterial flow may still occur, but venous outflow is often obstructed, causing diminished diastolic arterial flow on spectral Doppler examination157,158 (Fig. 21-29). If spontaneous detorsion occurs, flow within the affected testis may be normal, or it may be increased and mimic orchitis.159 Spontaneous detorsion rarely occurs and leaves a segmental testicular infarction.74,75 Segmental testicular infarction may also occur with Henoch-Schönlein purpura or with orchitis (see Fig. 21-14). Orchitis may also cause global ischemia of the testis and mimic torsion.159

868   PART III  ■  Small Parts, Carotid Artery, and Peripheral Vessel Sonography

A

B

C

D

E

F

G

H

I

FIGURE 21-29.  Torsion of spermatic cord and testis: spectrum of appearances. A to D, Acute torsion. Longitudinal power Doppler scans show A, no flow in the testis, and B, abnormal, transverse, and vertical orientation of the testis with no flow. C, After manual detorsion of case in B, longitudinal color Doppler scan shows the normal orientation of the testis with blood flow present. The testis has a striated appearance caused by the previous ischemia. D, Dual transverse gray-scale scan shows enlarged hypoechoic right testis resulting from torsion and skin thickening in the right hemiscrotum. E, Partial torsion. Longitudinal scan with spectral Doppler shows a high-resistance testicular arterial waveform with little diastolic flow because of venous occlusion; small, reactive hydrocele was found. F, After spontaneous detorsion of case in E, longitudinal scan with spectral Doppler shows return of diastolic flow. G, Torsion knot. Longitudinal scan with acute spermatic cord torsion shows the “torsion knot” complex of epididymis and spermatic cord. H, Acute torsion. Intraoperative photograph shows the twisted spermatic cord that gives the torsion knot appearance on sonograph. I, Subacute torsion (3 days of pain). Transverse power Doppler scan shows absent flow within the testis with surrounding hyperemia. (H from Winter TC. Ultrasonography of the scrotum. Appl Radiol 2002;31(3). H courtesy Drs. R.E. Berger, University of Washington, Seattle, and T.C. Winter, University of Wisconsin, Madison.)

In subacute or chronic torsion, color Doppler shows no flow in the testis and increased flow in the paratesticular tissues, including the epididymis-cord complex and dartos fascia (Fig. 21-29). Torsion of the testicular appendage is a common cause of acute scrotal pain and may mimic testicular

torsion clinically. Patients are rarely referred for imaging because the pain is usually not severe, and the twisted appendage may be evident clinically as the “blue dot” sign.160 The sonographic appearance of the twisted testicular appendage has been described as an avascular hypoechoic mass adjacent to a normally perfused testis

Chapter 21  ■  The Scrotum   869

and surrounded by an area of increased color Doppler perfusion.146 However, the twisted appendage may appear as an echogenic extratesticular mass situated between the head of the epididymis and the upper pole of the testis.161

Epididymitis and Epididymo-orchitis Epididymitis is the most common cause of acute scrotal pain in postpubertal men, causing 75% of all acute intrascrotal inflammatory processes. It usually results from a lower urinary tract infection and is less often hematogenous or traumatic in origin. The common causative organisms are Escherichia coli, Pseudomonas, and Klebsiella. Sexually transmitted organisms causing urethritis, such as gonococci and chlamydiae, are common causes of epididymitis in young men. Less frequently, epididymitis may be caused by tuberculosis, mumps, or syphilitic orchitis.162,163 The age of peak incidence is 40 to 50 years. Typically, patients present with the insidious onset of pain, which increases over 1 to 2 days. Fever, dysuria, and urethral discharge may also be present. In acute epididymitis, sonography characteristically shows thickening and enlargement of the epididymis, involving the tail initially and frequently spreading to the entire epididymis164 (Fig. 21-30, A and B). The echogenicity of the epididymis is usually decreased, and its echotexture is often coarse and heterogeneous, probably because of edema or hemorrhage, or both. Reactive hydrocele formation is common, and associated skin thickening may be seen. Color flow Doppler sonography usually shows increased blood flow in the epididymis or testis, or both, compared with the asymptomatic side165 (Fig. 21-30, C). Direct extension of epididymal inflammation to the testicle, called epididymo-orchitis, occurs in up to 20% of patients with acute epididymitis. Isolated orchitis may also occur. In such cases, increased blood flow is localized to the testis (Fig. 21-30, D and E; Video 21-6). Testicular involvement may be focal or diffuse. Characteristically, focal orchitis produces a hypoechoic area adjacent to an enlarged portion of the epididymis. Color Doppler shows increased flow in the hypoechoic area of the testis; increased flow in the tunica vasculosa may be visible as lines of color signal radiating from the mediastinum testis.166 These lines of color correspond to septal accentuation that is visible as hypoechoic bands on gray-scale sonography (Fig. 21-30, H and I). Spectral Doppler shows increased diastolic flow in uncomplicated orchitis (Fig. 21-31, A). If left untreated, the entire testicle may become involved, appearing hypoechoic and enlarged. As pressure in the testis increases from edema, venous infarction with hemorrhage may occur, appearing hyperechoic initially and hypoechoic later (Fig. 21-30).166 Ischemia and subsequent infarction may occur when the vascularity of the testis is compromised by venous occlu-

sion in the epididymis and cord.167 When vascular disruption is severe, resulting in complete testicular infarction, the changes are indistinguishable from those seen in testicular torsion. Color Doppler sonography may show focal areas of reactive hyperemia and increased blood flow associated with relatively avascular areas of infarction in both the testis and the epididymis in patients with severe epididymo-orchitis. Diastolic flow reversal in the arterial waveforms of the testis is an ominous finding, associated with testicular infarction in severe epididymo-orchitis168 (Fig. 21-31, B). In addition to infarction, other complications of acute epididymo-orchitis include abscess and pyocele (see Figs. 21-13 and 21-19, F). Chronic changes may be seen in the epididymis or testis from clinically resolved epididymo-orchitis. Swelling of the epididymis may persist and appear as a heterogeneous mass on sonography (see Fig. 21-24, A). The testis may have a persistent, striated appearance of septal accentuation from fibrosis (Figs. 21-32 and 21-33). This striated appearance of the testis is nonspecific and may also be seen after ischemia from torsion or during a hernia repair.166,169 A similar heterogeneous appearance in the testis may be seen in elderly patients because of seminiferous tubule atrophy and sclerosis.170 Focal areas of infarction in the testis may persist as wedge-shaped or cone-shaped hypoechoic areas or may appear as hyperechoic scars.166 If complete infarction of the testis has occurred because of epididymoorchitis, the testis may become small, with a hypoechoic or heterogeneous echotexture.

Fournier Gangrene Fournier gangrene is a necrotizing fasciitis of the perineum occurring most frequently in men age 50 to 70 years who are debilitated or who have diabetes mellitus.131 Multiple organisms are usually involved, including Klebsiella, Streptococcus, Proteus, and Staphylococcus. Surgical debridement of devitalized tissue is usually required, and morbidity and mortality are high without prompt treatment. Ultrasound may be helpful in diagnosis by showing scrotal wall thickening containing gas.

TRAUMA Prompt diagnosis of a ruptured testis is crucial because of the direct relationship between early surgical intervention and testicular salvageability. Approximately 90% of ruptured testicles can be saved if surgery is performed within the first 72 hours, whereas only 45% may be salvaged after 72 hours.171 Clinical diagnosis is often impossible because of marked scrotal pain and swelling, and sonography can be valuable in the assessment of tunica albuginea integrity and the extent of testicular hematoma.81,171-173 Sonographic features include focal areas of altered testicular

870   PART III  ■  Small Parts, Carotid Artery, and Peripheral Vessel Sonography

A

B

D

C E FIGURE 21-30.  Epididymo-orchitis, epididymitis, and orchitis: spectrum of appearances. A and B, Acute epididymitis. Longitudinal gray-scale and color Doppler images show enlargement and a heterogeneous echotexture of tail of the epididymis, with greatly increased flow in tail of the epididymis and minimally increased flow in the adjacent testis. C, Acute epididymo-orchitis. Longitudinal color Doppler scan shows increased flow in the epididymis and testis. D and E, Acute orchitis. Longitudinal dual-image gray-scale and color Doppler images show that right testis is hypoechoic and has greatly increased flow.

echogenicity, corresponding to areas of hemorrhage or infarction, and hematocele formation in 33% of patients. A discrete fracture plane is rarely identified (Fig. 21-34, B). A visibly intact tunica albuginea should exclude rupture, but testicular hematoma may obscure the tunica81 (Fig. 21-34, A). Tunical disruption associated with extrusion of the seminiferous tubules is specific for rupture (Fig. 21-34, E; Video 21-7). However, the sensitivity of the diagnosis of rupture based on tunical dis-

ruption alone is only 50%. Heterogeneity of the testis with associated testicular contour irregularity may be helpful in making the diagnosis of rupture.81,172,173 Although not specific for a ruptured testicle, these features may suggest the diagnosis in the appropriate clinical setting, prompting immediate surgical exploration. Color Doppler imaging can be helpful because rupture of the tunica albuginea is almost always associated with disruption of the tunica vasculosa and loss of

Chapter 21  ■  The Scrotum   871

F

G

H

I

FIGURE 21-30, cont’d. F, Longitudinal gray-scale scan with 3 weeks of epididymo-orchitis unresolved with antibiotic therapy shows hypoechoic areas in the testis and an enlarged heterogeneous tail of the epididymis. G, Color Doppler image shows increased flow in the testis and epididymis with an area of decreased flow due to ischemia (arrow). H and I, Acute orchitis. Longitudinal gray-scale and color Doppler images show hypoechoic bands caused by septal accentuation from edema and increased vascularity of the testis.

blood supply to part or all of the testis81 (Fig. 21-34, C ). A complex intrascrotal hematoma may be difficult to distinguish from testicular rupture.174 Patients with large, intrascrotal hematomas or hematoceles will often undergo surgical exploration because it is difficult to exclude rupture sonographically in the presence of surrounding complex fluid.81 Sonography can also be used to discern the severity of scrotal trauma resulting from bullet wounds, and foreign bodies can be localized.175 A careful gray-scale and color flow Doppler evaluation of the epididymis should be performed in all examinations done for blunt trauma. Traumatic epididymitis may be an isolated finding that should not be confused with an infectious process.176

CRYPTORCHIDISM The testes normally begin their descent through the inguinal canal into the scrotal sac at about 36 weeks of gestation. The gubernaculum testis is a fibromuscular structure that extends from the inferior pole of the testis to the scrotum and guides the testis in its descent, which normally has been completed at birth. Undescended testis is one of the most common genitourinary anomalies in male infants. At birth, 3.5% of male infants weighing more than 2500 g have an undescended testis; 10% to 25% of these cases are bilateral. This figure decreases to 0.8% by age 1 year because the testes des­ cend spontaneously in most infants. The incidence of

872   PART III  ■  Small Parts, Carotid Artery, and Peripheral Vessel Sonography

A

B

FIGURE 21-31.  Spectral Doppler changes in orchitis. A, Uncomplicated orchitis. Longitudinal scan with spectral Doppler tracing shows increased diastolic flow in the testis. B, Orchitis with venous compromise. Longitudinal scan with spectral Doppler tracing in more severe orchitis shows reversal of flow in diastole caused by edema impeding venous flow.

FIGURE 21-32.  Heterogeneous “striped” testis. Transverse dual image shows heterogeneity in the right testis with marked septal accentuation from previous orchitis. This appearance may also be seen after ischemia.

FIGURE 21-33.  Fibrosis of testis after orchitis. Pathologic specimen of testis shows linear bands of fibrosis (white areas) caused by previous severe orchitis. A similar “end stage” testis could have this appearance due to ischemia.

Chapter 21  ■  The Scrotum   873

H H A

D

B

C

E

F

FIGURE 21-34.  Testicular trauma: spectrum of appearances. A, Hematoma. Longitudinal image shows hematoma (arrows) on the anterior surface of the testis. Tunica intact at surgery. B, Fracture of testis. Transverse scan shows a heterogeneous testicle with a linear band (arrows) indicating a fracture. H, Testicular hematoma. C, Tunical tear. Longitudinal color Doppler image shows contour irregularity of the testis with disruption of the tunica (arrow). Extruded testis parenchyma shows no color flow. D, Same case as C. Photograph during surgery shows tunical tear in the exposed right testis. E, Rupture of testis. Longitudinal image shows rupture of the testis with extrusion of seminiferous tubules (arrow). F, In same case as E, photograph during surgery shows a tear in the tunica inferiorly with extrusion of seminiferous tubules.

undescended testes increases to 30% in premature infants, approaching 100% in neonates who weigh less than 1 kg at birth. Complete descent is necessary for full testicular maturation.177,178 Malpositioned testes may be located anywhere along the pathway of descent from the retroperitoneum to the scrotum. Most (80%) undescended testes are palpable, lying at or below the level of the inguinal canal. Anorchia occurs in 4% of the remaining patients with impalpable testes.178 Localization of the undescended testis is important for the prevention of two potential complications of cryptorchidism: infertility and cancer. The undescended testis is more likely to undergo malignant change than the normally descended testis.1 The most common malignancy is seminoma. The risk of malignancy is increased in both the undescended testis after orchiopexy and the normally descended testis. Therefore, careful serial examinations of both testes are essential. Sonographically, the undescended testis is often smaller and slightly less echogenic than the contralateral, normally descended testis (Fig. 21-35). A large lymph node or the pars infravaginalis gubernaculi (PIG), which is the distal bulbous segment of the gubernaculum testis, can be mistaken for the testis. After completion of testicular descent, the PIG and the gubernaculum normally

FIGURE 21-35.  Testis in inguinal canal. Longitudinal scan shows an elongated, ovoid, undescended testis.

atrophy. If the testis remains undescended, both structures persist. The PIG is located distal to the undescended testis, usually in the scrotum, but it may be found in the inguinal cord. Sonographically, the PIG is a hypoechoic, cordlike structure of echogenicity similar to the testis, with the gubernaculum leading to it.179 Sonography is often used in the initial evaluation of cryptorchidism, although the value of this has been

874   PART III  ■  Small Parts, Carotid Artery, and Peripheral Vessel Sonography

questioned because it is insensitive in detecting high intra-abdominal testes.180 MRI has also been used in cryptorchidism because it is more sensitive than ultrasound in detecting undescended testes in the retroperitoneum.181,182 Nonvisualization of an undescended testis on sonography or MRI does not exclude its presence, and therefore laparoscopy or surgical exploration should be performed if clinically indicated.

Acknowledgment Frank Thornton, MD, assisted in gathering images.

References Anatomy 1. Krone KD, Carroll BA. Scrotal ultrasound. Radiol Clin North Am 1985;23:121-139. 2. Trainer TD. Histology of the normal testis. Am J Surg Pathol 1987;11:797-809. 3. Johnson KA, Dewbury KC. Ultrasound imaging of the appendix testis and appendix epididymis. Clin Radiol 1996;51:335-337. 4. Sellars ME, Sidhu PS. Ultrasound appearances of the testicular appendages: pictorial review. Eur Radiol 2003;13:127-135. 5. Black JA, Patel A. Sonography of the normal extratesticular space. AJR Am J Roentgenol 1996;167:503-506. 6. Allen TD. Disorders of the male external genitalia. In: Kelalis PP, King LR, editors. Clinical pediatric urology. Philadelphia: Saunders; 1976. p. 636-668. 7. Middleton WD, Bell MW. Analysis of intratesticular arterial anatomy with emphasis on transmediastinal arteries. Radiology 1993;189:157-160. 8. Fakhry J, Khoury A, Barakat K. The hypoechoic band: a normal finding on testicular sonography. AJR Am J Roentgenol 1989;153: 321-323. 9. Bushby LH, Sellars ME, Sidhu PS. The “two-tone” testis: spectrum of ultrasound appearances. Clin Radiol 2007;62:1119-1123. 10. Middleton WD, Thorne DA, Melson GL. Color Doppler ultrasound of the normal testis. AJR Am J Roentgenol 1989;152:293-297. 11. Gooding GA. Sonography of the spermatic cord. AJR Am J Roentgenol 1988;151:721-724. Scrotal Masses 12. Rifkin MD, Kurtz AB, Pasto ME, Goldberg BB. Diagnostic capabilities of high-resolution scrotal ultrasonography: prospective evaluation. J Ultrasound Med 1985;4:13-19. 13. Carroll BA, Gross DM. High-frequency scrotal sonography. AJR Am J Roentgenol 1983;140:511-515. 14. Grantham JG, Charboneau JW, James EM, et al. Testicular neoplasms: 29 tumors studied by high-resolution ultrasound. Radiology 1985;157:775-780. 15. Kirschling RJ, Kvols LK, Charboneau JW, et al. High-resolution ultrasonographic and pathologic abnormalities of germ cell tumors in patients with clinically normal testes. Mayo Clin Proc 1983;58: 648-653. 16. Javadpour N. Principles and management of testicular cancer. New York, Thieme; 1986. 17. Damjanov I. Tumors of the testis and epididymis. In: Murphy WM, editor. Urological pathology. 2nd ed. Philadelphia: Saunders; 1997. p. 342-400. 18. Talerman A, Roth LM. Pathology of the testis and its adnexa. In: Talerman A, Roth LM, editors. Germ cell tumors. New York: Churchill Livingstone; 1986. 19. Jacobsen GK, Talerman A. Atlas of germ cell tumors. J Pathol 1989; 160:86-87. 20. Schwerk WB, Schwerk WN, Rodeck G. Testicular tumors: prospective analysis of real-time ultrasound patterns and abdominal staging. Radiology 1987;164:369-374. 21. Ulbright TM. Germ cell neoplasms of the testis. Am J Surg Pathol 1993;17:1075-1091.

22. Mostofi FK, Price EB. Tumors of the male genital system. In: Atlas of tumor pathology. Washington, DC: Armed Forces Institute of Pathology; 1973. 23. Emory TH, Charboneau JW, Randall RV, et al. Occult testicular interstitial-cell tumor in a patient with gynecomastia: ultrasonic detection. Radiology 1984;151:474. 24. Maizlin ZV, Belenky A, Kunichezky M, et al. Leydig cell tumors of the testis: gray scale and color Doppler sonographic appearance. J Ultrasound Med 2004;23:959-964. 25. Gabrilove JL, Freiberg EK, Leiter E, Nicolis GL. Feminizing and non-feminizing Sertoli cell tumors. J Urol 1980;124:757767. 26. Young S, Gooneratne S, Straus Jr FH, et al. Feminizing Sertoli cell tumors in boys with Peutz-Jeghers syndrome. Am J Surg Pathol 1995;19:50-58. 27. Tasu JP, Faye N, Eschwege P, et al. Imaging of burned-out testis tumor: five new cases and review of the literature. J Ultrasound Med 2003;22:515-521. 28. Mindrup SR, Konety BR. Testicular recurrence from “primary” retroperitoneal germ cell tumor. Urology 2004;64:1031. 29. Shawker TH, Javadpour N, O’Leary T, et al. Ultrasonographic detection of “burned-out” primary testicular germ cell tumors in clinically normal testes. J Ultrasound Med 1983;2:477-479. 30. Horstman WG, Melson GL, Middleton WD, Andriole GL. Testicular tumors: findings with color Doppler ultrasound. Radiology 1992;185:733-737. 31. Luker GD, Siegel MJ. Pediatric testicular tumors: evaluation with gray-scale and color Doppler ultrasound. Radiology 1994;191:561564. 32. Hindley RG, Chandra A, Saunders A, O’Brien TS. Impalpable testis cancer. BJU Int 2003;92:572-574. 33. Powell TM, Tarter TH. Management of nonpalpable incidental testicular masses. J Urol 2006;176:96-98; discussion 99. 34. Carmignani L, Gadda F, Gazzano G, et al. High incidence of benign testicular neoplasms diagnosed by ultrasound. J Urol 2003;170:17831786. 35. Carmignani L, Morabito A, Gadda F, et al. Prognostic parameters in adult impalpable ultrasonographic lesions of the testicle. J Urol 2005;174:1035-1038. 36. Horstman WG, Haluszka MM, Burkhard TK. Management of testicular masses incidentally discovered by ultrasound. J Urol 1994; 151:1263-1265. 37. Mostofi FK, Sobin LH. Histological typing of testis tumours. In: International histological classification of tumors of the testes. Geneva: World Health Organization; 1977. 38. Mazzu D, Jeffrey Jr RB, Ralls PW. Lymphoma and leukemia involving the testicles: findings on gray-scale and color Doppler sonography. AJR Am J Roentgenol 1995;164:645-647. 39. Rayor RA, Scheible W, Brock WA, Leopold GR. High-resolution ultrasonography in the diagnosis of testicular relapse in patients with acute lymphoblastic leukemia. J Urol 1982;128:602-603. 40. Iizumi T, Shinohara S, Amemiya H, et al. Plasmacytoma of the testis. Urol Int 1995;55:218-221. 41. Grignon DJ, Shum DT, Hayman WP. Metastatic tumours of the testes. Can J Surg 1986;29:359-361. 42. Werth V, Yu G, Marshall FF. Nonlymphomatous metastatic tumor to the testis. J Urol 1982;127:142-144. 43. Hanash KA, Carney JA, Kelalis PP. Metastatic tumors to testicles: routes of metastasis. J Urol 1969;102:465-468. 44. Hamm B, Fobbe F, Loy V. Testicular cysts: differentiation with ultrasound and clinical findings. Radiology 1988;168:19-23. 45. Shergill IS, Thwaini A, Kapasi F, et al. Management of simple intratesticular cysts: a single-institution 11-year experience. Urology 2006;67:1266-1268. 46. Martinez-Berganza MT, Sarria L, Cozcolluela R, et al. Cysts of the tunica albuginea: sonographic appearance. AJR Am J Roentgenol 1998;170:183-185. 47. Dogra VS, Gottlieb RH, Rubens DJ, Liao L. Benign intratesticular cystic lesions: ultrasound features. Radiographics 2001;21 Spec No:273-281. 48. Warner KE, Noyes DT, Ross JS. Cysts of the tunica albuginea testis: a report of 3 cases with a review of the literature. J Urol 1984; 132:131-132. 49. Poster RB, Spirt BA, Tamsen A, Surya BV. Complex tunica albuginea cyst simulating an intratesticular lesion. Urol Radiol 1991;13: 129-132.

Chapter 21  ■  The Scrotum   875 50. Sudakoff GS, Quiroz F, Karcaaltincaba M, Foley WD. Scrotal ultrasonography with emphasis on the extratesticular space: anatomy, embryology, and pathology. Ultrasound Q 2002;18:255-273. 51. Takihara H, Valvo JR, Tokuhara M, Cockett AT. Intratesticular cysts. Urology 1982;20:80-82. 52. Tartar VM, Trambert MA, Balsara ZN, Mattrey RF. Tubular ectasia of the testicle: sonographic and MR imaging appearance. AJR Am J Roentgenol 1993;160:539-542. 53. Brown DL, Benson CB, Doherty FJ, et al. Cystic testicular mass caused by dilated rete testis: sonographic findings in 31 cases. AJR Am J Roentgenol 1992;158:1257-1259. 54. Weingarten BJ, Kellman GM, Middleton WD, Gross ML. Tubular ectasia within the mediastinum testis. J Ultrasound Med 1992;11: 349-353. 55. Older RA, Watson LR. Tubular ectasia of the rete testis: a benign condition with a sonographic appearance that may be misinterpreted as malignant. J Urol 1994;152:477-478. 56. Cho CS, Kosek J. Cystic dysplasia of the testis: sonographic and pathologic findings. Radiology 1985;156:777-778. 57. Keetch DW, McAlister WH, Manley CB, Dehner LP. Cystic dysplasia of the testis: sonographic features with pathologic correlation. Pediatr Radiol 1991;21:501-503. 58. Atchley JT, Dewbury KC. Ultrasound appearances of testicular epidermoid cysts. Clin Radiol 2000;55:493-502. 59. Sanderson AJ, Birch BR, Dewbury KC. Case report: multiple epidermoid cysts of the testes–the ultrasound appearances. Clin Radiol 1995;50:414-415. 60. Malvica RP. Epidermoid cyst of the testicle: an unusual sonographic finding. AJR Am J Roentgenol 1993;160:1047-1048. 61. Stein MM, Stein MW, Cohen BC, et al. Unusual sonographic appearance of an epidermoid cyst of the testis. J Ultrasound Med 1999;18:723-726. 62. Maizlin ZV, Belenky A, Baniel J, et al. Epidermoid cyst and teratoma of the testis: sonographic and histologic similarities. J Ultrasound Med 2005;24:1403-1409; quiz 1410-1411. 63. Eisenmenger M, Lang S, Donner G, et al. Epidermoid cysts of the testis: organ-preserving surgery following diagnosis by ultrasonography. Br J Urol 1993;72:955-957. 64. Cho JH, Chang JC, Park BH, et al. Sonographic and MR imaging findings of testicular epidermoid cysts [see comment]. AJR Am J Roentgenol 2002;178:743-748. 65. Langer JE, Ramchandani P, Siegelman ES, Banner MP. Epidermoid cysts of the testicle: sonographic and MR imaging features. AJR Am J Roentgenol 1999;173:1295-1299. 66. Hermansen MC, Chusid MJ, Sty JR. Bacterial epididymo-orchitis in children and adolescents. Clin Pediatr 1980;19:812-815. 67. Mevorach RA, Lerner RM, Dvoretsky PM, Rabinowitz R. Testicular abscess: diagnosis by ultrasonography. J Urol 1986;136:12131216. 68. Korn RL, Langer JE, Nisenbaum HL, Miller Jr WT, Cheung LP. Non-Hodgkin’s lymphoma mimicking a scrotal abscess in a patient with AIDS. Journal of Ultrasound in Medicine 1994;13:715-718. 69. Smith FJ, Bilbey JH, Filipenko JD, Goldenberg SL. Testicular pseudotumor in the acquired immunodeficiency syndrome. Urology 1995;45:535-537. 70. Wu VH, Dangman BC, Kaufman Jr RP. Sonographic appearance of acute testicular venous infarction in a patient with a hypercoagulable state. J Ultrasound Med 1995;14:57-59. 71. Bilagi P, Sriprasad S, Clarke JL, et al. Clinical and ultrasound features of segmental testicular infarction: six-year experience from a single centre. Eur Radiol 2007;17:2810-2818. 72. Flanagan JJ, Fowler RC. Testicular infarction mimicking tumour on scrotal ultrasound: a potential pitfall. Clin Radiol 1995;50:4950. 73. Einstein DM, Paushter DM, Singer AA, et al. Fibrotic lesions of the testicle: sonographic patterns mimicking malignancy. Urol Radiol 1992;14:205-210. 74. Ledwidge ME, Lee DK, Winter 3rd TC, et al. Sonographic diagnosis of superior hemispheric testicular infarction.[see comment]. AJR Am J Roentgenol 2002;179:775-776. 75. Sriprasad S, Kooiman GG, Muir GH, Sidhu PS. Acute segmental testicular infarction: differentiation from tumour using highfrequency colour Doppler ultrasound. Br J Radiol 2001;74:965967. 76. Carmody JP, Sharma OP. Intrascrotal sarcoidosis: case reports and review. Sarcoidosis Vasc Diffuse Lung Dis 1996;13:129-134.

77. Winter 3rd TC, Keener TS, Mack LA. Sonographic appearance of testicular sarcoid. J Ultrasound Med 1995;14:153-156. 78. Eraso CE, Vrachliotis TG, Cunningham JJ. Sonographic findings in testicular sarcoidosis simulating malignant nodule. J Clin Ultrasound 1999;27:81-83. 79. Avila NA, Premkumar A, Shawker TH, et al. Testicular adrenal rest tissue in congenital adrenal hyperplasia: findings at Gray-scale and color Doppler ultrasound. Radiology 1996;198:99-104. 80. Vanzulli A, DelMaschio A, Paesano P, et al. Testicular masses in association with adrenogenital syndrome: ultrasound findings. Radiology 1992;183:425-429. 81. Bhatt S, Dogra VS. Role of ultrasound in testicular and scrotal trauma. Radiographics 2008;28:1617-1629. 82. Gierke CL, King BF, Bostwick DG, et al. Large-cell calcifying Sertoli cell tumor of the testis: appearance at sonography. AJR Am J Roentgenol 1994;163:373-375. 83. Vegni-Talluri M, Bigliardi E, Vanni MG, Tota G. Testicular microliths: their origin and structure. J Urol 1980;124:105-107. 84. Breger RC, Passarge E, McAdams AJ. Testicular intratubular bodies. J Clin Endocrinol Metab 1965;25:1340-1346. 85. Middleton WD, Teefey SA, Santillan CS. Testicular microlithiasis: prospective analysis of prevalence and associated tumor. Radiology 2002;224:425-428. 86. Kim B, Winter 3rd TC, Ryu JA. Testicular microlithiasis: clinical significance and review of the literature. Eur Radiol 2003;13: 2567-2576. 87. Nistal M, Paniagua R, Diez-Pardo JA. Testicular microlithiasis in 2 children with bilateral cryptorchidism. J Urol 1979;121:535-537. 88. Janzen DL, Mathieson JR, Marsh JI, et al. Testicular microlithiasis: sonographic and clinical features [see comment]. AJR Am J Roentgenol 1992;158:1057-1060. 89. Backus ML, Mack LA, Middleton WD, et al. Testicular microlithiasis: imaging appearances and pathologic correlation. Radiology 1994;192:781-785. 90. Patel MD, Olcott EW, Kerschmann RL, et al. Sonographically detected testicular microlithiasis and testicular carcinoma. J Clin Ultrasound 1993;21:447-452. 91. Cast JE, Nelson WM, Early AS, et al. Testicular microlithiasis: prevalence and tumor risk in a population referred for scrotal sonography. AJR Am J Roentgenol 2000;175:1703-1706. 92. Bennett HF, Middleton WD, Bullock AD, Teefey SA. Testicular microlithiasis: ultrasound follow-up. Radiology 2001;218:359363. 93. Bach AM, Hann LE, Hadar O, et al. Testicular microlithiasis: what is its association with testicular cancer [see comment]? Radiology 2001;220:70-75. 94. Frush DP, Kliewer MA, Madden JF. Testicular microlithiasis and subsequent development of metastatic germ cell tumor. AJR Am J Roentgenol 1996;167:889-890. 95. Smith WS, Brammer HM, Henry M, Frazier H. Testicular microlithiasis: sonographic features with pathologic correlation. AJR Am J Roentgenol 1991;157:1003-1004. 96. McEniff N, Doherty F, Katz J, Schrager CA, Klauber G. Yolk sac tumor of the testis discovered on a routine annual sonogram in a boy with testicular microlithiasis. AJR Am J Roentgenol 1995;164: 971-972. 97. Miller FN, Sidhu PS. Does testicular microlithiasis matter? A review [see comment]. Clin Radiol 2002;57:883-890. 98. Quane LK, Kidney DD. Testicular microlithiasis in a patient with a mediastinal germ cell tumour [see comment]. Clin Radiol 2000; 55:642-644. 99. Dagash H, Mackinnon EA. Testicular microlithiasis: what does it mean clinically? BJU Int 2007;99:157-160. 100. Lam DL, Gerscovich EO, Kuo MC, McGahan JP. Testicular microlithiasis: our experience of 10 years. J Ultrasound Med 2007;26: 867-873. 101. Ringdhal E, Claybrook K, Teague JL, et al. Testicular microlithiasis and its relation to testicular cancer on ultrasound findings of symptomatic men. J Urol 2004;172:1904-1906. 102. Sakamoto H, Shichizyou T, Saito K, et al. Testicular microlithiasis identified ultrasonographically in Japanese adult patients: prevalence and associated conditions. Urology 2006;68:636-641. 103. Konstantinos S, Alevizos A, Anargiros M, et al. Association between testicular microlithiasis, testicular cancer, cryptorchidism and history of ascending testis. Int Braz J Urol 2006;32:434-438; discussion 439.

876   PART III  ■  Small Parts, Carotid Artery, and Peripheral Vessel Sonography 104. Serter S, Gumos B, Unlu M, et al. Prevalence of testicular microlithiasis in an asymptomatic population. Scand J Urol Nephrol 2006; 40:212-214. 105. Linkowski GD, Avellone A, Gooding GA. Scrotal calculi: sonographic detection. Radiology 1985;156:484. 106. Rathaus V, Konen O, Shapiro M, et al. Ultrasound features of spermatic cord hydrocele in children. Br J Radiol 2001;74:818820. 107. Nye PJ, Prati Jr RC. Idiopathic hydrocele and absent testicular diastolic flow. J Clin Ultrasound 1997;25:43-46. 108. Worthy L, Miller EI, Chinn DH. Evaluation of extratesticular findings in scrotal neoplasms. J Ultrasound Med 1986;5:261263. 109. Gooding GA, Leonhardt WC, Marshall G, et al. Cholesterol crystals in hydroceles: sonographic detection and possible significance. AJR Am J Roentgenol 1997;169:527-529. 110. Cunningham JJ. Sonographic findings in clinically unsuspected acute and chronic scrotal hematoceles. AJR Am J Roentgenol 1983; 140:749-752. 111. Beddy P, Geoghegan T, Browne RF, Torreggiani WC. Testicular varicoceles. Clin Radiol 2005;60:1248-1255. 112. Zucchi A, Mearini L, Mearini E, et al. Varicocele and fertility: relationship between testicular volume and seminal parameters before and after treatment. J Androl 2006;27:548-551. 113. Gonda RL Jr, Karo JJ, Forte RA, O’Donnell KT. Diagnosis of subclinical varicocele in infertility. AJR Am J Roentgenol 1987; 148:71-75. 114. Graif M, Hauser R, Hirshebein A, et al. Varicocele and the testicularrenal venous route: hemodynamic Doppler sonographic investigation. J Ultrasound Med 2000;19:627-631. 115. Tetreau R, Julian P, Lyonnet D, Rouviere O. Intratesticular varicocele: an easy diagnosis but unclear physiopathologic characteristics. J Ultrasound Med 2007;26:1767-1773. 116. Atasoy C, Fitoz S. Gray-scale and color Doppler sonographic findings in intratesticular varicocele. J Clin Ultrasound 2001;29: 369-373. 117. Bhosale PR, Patnana M, Viswanathan C, Szklaruk J. The inguinal canal: anatomy and imaging features of common and uncommon masses. Radiographics 2008;28:819-835; quiz 913. 118. Sung T, Riedlinger WF, Diamond DA, Chow JS. Solid extratesticular masses in children: radiographic and pathologic correlation. AJR Am J Roentgenol 2006;186:483-490. 119. Pavone-MacAluso M, Smith PH, Bagshaw MA. Testicular cancer and other tumors of the genitourinary tract. New York: Plenum Press; 1985. 120. Alleman WG, Gorman B, King BF, et al. Benign and malignant epididymal masses evaluated with scrotal sonography: clinical and pathologic review of 85 patients. J Ultrasound Med 2008;27: 1195-1202. 121. Yang DM, Kim SH, Kim HN, et al. Differential diagnosis of focal epididymal lesions with gray scale sonographic, color Doppler sonographic, and clinical features. J Ultrasound Med 2003;22:135-142; quiz 143-144. 122. Smallman LA, Odedra JK. Primary carcinoma of sigmoid colon metastasizing to epididymis. Urology 1984;23:598-599. 123. Wachtel TL, Mehan DJ. Metastatic tumors of the epididymis. J Urol 1970;103:624-627. 124. Oh C, Nisenbaum HL, Langer J, et al. Sonographic demonstration, including color Doppler imaging, of recurrent sperm granuloma. J Ultrasound Med 2000;19:333-335. 125. Krainik A, Sarrazin JL, Camparo P, et al. Fibrous pseudotumor of the epididymis: imaging and pathologic correlation. Eur Radiol 2000;10:1636-1638. 126. Al-Otaibi L, Whitman GJ, Chew FS. Fibrous pseudotumor of the epididymis. AJR Am J Roentgenol 1997;168:1586. 127. Oliva E, Young RH. Paratesticular tumor-like lesions. Semin Diagn Pathol 2000;17:340-358. 128. Leung ML, Gooding GA, Williams RD. High-resolution sonography of scrotal contents in asymptomatic subjects. AJR Am J Roentgenol 1984;143:161-164. 129. Reddy NM, Gerscovich EO, Jain KA, et al. Vasectomy-related changes on sonographic examination of the scrotum. J Clin Ultrasound 2004;32:394-398. 130. Ishigami K, Abu-Yousef MM, El-Zein Y. Tubular ectasia of the epididymis: a sign of postvasectomy status. J Clin Ultrasound 2005; 33:447-451.

131. Stewart VR, Sidhu PS. The testis: the unusual, the rare and the bizarre. Clin Radiol 2007;62:289-302. Acute Scrotal Pain 132. Mueller DL, Amundson GM, Rubin SZ, Wesenberg RL. Acute scrotal abnormalities in children: diagnosis by combined sonography and scintigraphy. AJR Am J Roentgenol 1988;150:643-646. 133. Watanabe Y, Dohke M, Ohkubo K, et al. Scrotal disorders: evaluation of testicular enhancement patterns at dynamic contrastenhanced subtraction MR imaging [see comment]. Radiology 2000;217:219-227. 134. Hricak H, Lue T, Filly RA, et al. Experimental study of the sonographic diagnosis of testicular torsion. J Ultrasound Med 1983;2:349-356. 135. Pillai SB, Besner GE. Pediatric testicular problems. Pediatr Clin North Am 1998;45:813-830. 136. Paltiel HJ. Sonography of pediatric scrotal emergencies. Ultrasound Q 2000;16:53-71. 137. Prando D. Torsion of the spermatic cord: sonographic diagnosis. Ultrasound Q 2002;18:41-57. 138. Sidhu PS. Clinical and imaging features of testicular torsion: role of ultrasound. Clin Radiol 1999;54:343-352. 139. Middleton WD, Melson GL. Testicular ischemia: color Doppler sonographic findings in five patients. AJR Am J Roentgenol 1989; 152:1237-1239. 140. Chinn DH, Miller EI. Generalized testicular hyperechogenicity in acute testicular torsion. J Ultrasound Med 1985;4:495-496. 141. Winter TC 3rd. Ultrasonography of the scrotum. App Radiol 2002;31. 142. Middleton WD, Middleton MA, Dierks M, et al. Sonographic prediction of viability in testicular torsion: preliminary observations. J Ultrasound Med 1997;16:23-27; quiz 29-30. 143. Vijayaraghavan SB. Sonographic differential diagnosis of acute scrotum: real-time whirlpool sign, a key sign of torsion. J Ultrasound Med 2006;25:563-574. 144. Trambert MA, Mattrey RF, Levine D, Berthoty DP. Subacute scrotal pain: evaluation of torsion versus epididymitis with MR imaging. Radiology 1990;175:53-56. 145. Vick CW, Bird K, Rosenfield AT, et al. Extratesticular hemorrhage associated with torsion of the spermatic cord: sonographic demonstration. Radiology 1986;158:401-404. 146. Lerner RM, Mevorach RA, Hulbert WC, Rabinowitz R. Color Doppler ultrasound in the evaluation of acute scrotal disease. Radiology 1990;176:355-358. 147. Fitzgerald SW, Erickson S, DeWire DM, et al. Color Doppler sonography in the evaluation of the adult acute scrotum [see comment]. J Ultrasound Med 1992;11:543-548. 148. Barth RA, Shortliffe LD. Normal pediatric testis: comparison of power Doppler and color Doppler ultrasound in the detection of blood flow. Radiology 1997;204:389-393. 149. Bader TR, Kammerhuber F, Herneth AM. Testicular blood flow in boys as assessed at color Doppler and power Doppler sonography. Radiology 1997;202:559-564; erratum 203:580. 150. Oley BD, Frush DP, Babcock DS, et al. Acute testicular torsion: comparison of unenhanced and contrast-enhanced power Doppler ultrasound, color Doppler ultrasound, and radionuclide imaging. Radiology 1996;199:441-446. 151. Luker GD, Siegel MJ. Scrotal US in pediatric patients: comparison of power and standard color Doppler ultrasound. Radiology 1996; 198:381-385. 152. Albrecht T, Lotzof K, Hussain HK, Shedden D, Cosgrove DO, de Bruyn R. Power Doppler US of the normal prepubertal testis: does it live up to its promises? Radiology 1997;203:227-231. 153. Lee Jr FT, Winter DB, Madsen FA, et al. Conventional color Doppler velocity sonography versus color Doppler energy sonography for the diagnosis of acute experimental torsion of the spermatic cord. AJR Am J Roentgenol 1996;167:785-790. 154. Burks DD, Markey BJ, Burkhard TK, et al. Suspected testicular torsion and ischemia: evaluation with color Doppler sonography. Radiology 1990;175:815-821. 155. Atkinson Jr GO, Patrick LE, Ball Jr TI, et al. The normal and abnormal scrotum in children: evaluation with color Doppler sonography. AJR Am J Roentgenol 1992;158:613-617. 156. Bude RO, Kennelly MJ, Adler RS, Rubin JM. Nonpulsatile arterial waveforms: observations during graded testicular torsion in rats. Acad Radiol 1995;2:879-882.

Chapter 21  ■  The Scrotum   877 157. Dogra VS, Rubens DJ, Gottlieb RH, Bhatt S. Torsion and beyond: new twists in spectral Doppler evaluation of the scrotum. J Ultrasound Med 2004;23:1077-1085. 158. Sanelli PC, Burke BJ, Lee L. Color and spectral Doppler sonography of partial torsion of the spermatic cord. AJR Am J Roentgenol 1999;172:49-51. 159. Alcantara AL, Sethi Y. Imaging of testicular torsion and epidi­ dymitis/orchitis: diagnosis and pitfalls. Emerg Radiol 1998;5:394402. 160. Dresner ML. Torsed appendage: diagnosis and management—blue dot sign. Urology 1973;1:63-66. 161. Hesser U, Rosenborg M, Gierup J, et al. Gray-scale sonography in torsion of the testicular appendages. Pediatr Radiol 1993;23:529532. 162. Chung JJ, Kim MJ, Lee T, et al. Sonographic findings in tuberculous epididymitis and epididymo-orchitis. J Clin Ultrasound 1997;25: 390-394. 163. Basekim CC, Kizilkaya E, Pekkafali Z, et al. Mumps epididymoorchitis: sonography and color Doppler sonographic findings. Abdom Imaging 2000;25:322-325. 164. Gondos B, Wong TW. Non-neoplastic diseases of the testis and epididymis. In: Murphy WM, editor. Urological pathology. 2nd ed. Philadelphia: Saunders; 1997. p. 277-341. 165. Horstman WG, Middleton WD, Melson GL. Scrotal inflammatory disease: color Doppler ultrasound findings. Radiology 1991;179: 55-59. 166. Cook JL, Dewbury K. The changes seen on high-resolution ultrasound in orchitis. Clin Radiol 2000;55:13-18. 167. Hourihane DO. Infected infarcts of the testis: a study of 18 cases preceded by pyogenic epididymoorchitis. J Clin Pathol 1970;23: 668-675. 168. Sanders LM, Haber S, Dembner A, Aquino A. Significance of reversal of diastolic flow in the acute scrotum. J Ultrasound Med 1994;13:137-139. 169. Casalino DD, Kim R. Clinical importance of a unilateral striated pattern seen on sonography of the testicle. AJR Am J Roentgenol 2002;178:927-930. 170. Harris RD, Chouteau C, Partrick M, Schned A. Prevalence and significance of heterogeneous testes revealed on sonography: ex vivo

sonographic-pathologic correlation. AJR Am J Roentgenol 2000; 175:347-352. Trauma 171. Jeffrey RB, Laing FC, Hricak H, McAninch JW. Sonography of testicular trauma. AJR Am J Roentgenol 1983;141:993-995. 172. Kim SH, Park S, Choi SH, et al. Significant predictors for determination of testicular rupture on sonography: a prospective study. J Ultrasound Med 2007;26:1649-1655. 173. Buckley JC, McAninch JW. Use of ultrasonography for the diagnosis of testicular injuries in blunt scrotal trauma. J Urol 2006;175: 175-178. 174. Cohen HL, Shapiro ML, Haller JO, Glassberg K. Sonography of intrascrotal hematomas simulating testicular rupture in adolescents. Pediatr Radiol 1992;22:296-297. 175. Learch TJ, Hansch LP, Ralls PW. Sonography in patients with gunshot wounds of the scrotum: imaging findings and their value. AJR Am J Roentgenol 1995;165:879-883. 176. Gordon LM, Stein SM, Ralls PW. Traumatic epididymitis: evaluation with color Doppler sonography. AJR Am J Roentgenol 1996; 166:1323-1325. Cryptorchidism 177. Elder JS. Cryptorchidism: isolated and associated with other genitourinary defects. Pediatr Clin North Am 1987;34:1033-1053. 178. Harrison JH, Gittes RF, Stamey TA, et al. Campbell’s urology. 4th ed. Philadelphia: Saunders; 1979. 179. Rosenfield AT, Blair DN, McCarthy S, et al. The pars infravaginalis gubernaculi: importance in the identification of the undescended testis. Society of Uroradiology Award paper. AJR Am J Roentgenol 1989;153:775-778. 180. Friedland GW, Chang P. The role of imaging in the management of the impalpable undescended testis. AJR Am J Roentgenol 1988; 151:1107-1111. 181. Fritzsche PJ, Hricak H, Kogan BA, et al. Undescended testis: value of MR imaging. Radiology 1987;164:169-173. 182. Kier R, McCarthy S, Rosenfield AT, et al. Nonpalpable testes in young boys: evaluation with MR imaging. Radiology 1988;169: 429-433.

CHAPTER 22 

The Rotator Cuff Marnix T. van Holsbeeck, Dzung Vu, and J. Antonio Bouffard

Chapter Outline CLINICAL CONSIDERATIONS TECHNICAL CONSIDERATIONS ANATOMY AND SONOGRAPHIC TECHNIQUE THE NORMAL CUFF The Adolescent Cuff Age-Related Changes

PREOPERATIVE APPEARANCES Criteria of Rotator Cuff Tears Nonvisualization of the Cuff Focal Nonvisualization of the Cuff Discontinuity in the Cuff Focal Abnormal Echogenicity

Associated Findings

Subdeltoid Bursal Effusion Joint Effusion

Shoulder pain has many causes. Tendinitis, rotator

cuff strain, and partial-thickness or full-thickness tear may cause pain and weakness on elevation of the arm.1 The pain in rotator cuff disease is often worse at night and may keep the patient awake. Underlying these symptoms in many patients over 40 years of age is failure of the rotator cuff fibers.2 The supraspinatus tendon fibers typically fail first. The subscapularis and infraspinatus tendons, two other tendons of the rotator cuff, fail when the tear extends. The teres minor, the fourth component of the rotator cuff, is rarely affected. Calcific tendinitis, cervical radiculopathy, and acromioclavicular arthritis may mimic rotator cuff pathology. Contrast arthrography has long been the premier radiologic examination used to diagnose full-thickness tears of the rotator cuff.3 Two competing noninvasive imaging techniques, ultrasound and magnetic resonance imaging (MRI), are taking over the role of arthrography. High-resolution real-time ultrasound has been shown to be a cost-effective means of examining the rotator cuff.4-9 Ultrasound is the modality of choice in our institution. In the last 15 years, we performed more than 40,000 shoulder ultrasound studies.

CLINICAL CONSIDERATIONS Rotator cuff fiber failure is the most common cause of shoulder pain and dysfunction in patients older than 40.1 Epidemiologic studies by Codman, DePalma, and others have demonstrated that the frequency of rotator cuff fiber failure increases with age.10-12 This aging of tendons 878

Concave Subdeltoid Fat Contour Bone Surface Irregularity Tear Size and Muscle Atrophy

Pathology of Rotator Cuff Interval POSTOPERATIVE APPEARANCES Recurrent Tear PITFALLS IN INTERPRETATION ROTATOR CUFF CALCIFICATIONS

has been shown in imaging studies as well.13-16 The earliest changes are often located in the substance of the tendon, resulting in so-called delamination of the cuff. Fiber failure is a step-by-step process from partial-thickness tear, almost always first in the supraspinatus, to massive tears involving multiple cuff tendons. Rotator cuff tear may occur insidiously and, in fact, may be unnoticed by the patient, a process termed by some as creeping tendon ruptures.17 Asymptomatic tears affect up to 30% of the population over age 60.13 When a larger group of fibers fails at once, the shoulder demonstrates pain at rest and accentuation of pain on use of the rotator cuff (e.g., extension, abduction, or external rotation). When even greater numbers of fibers fail at one time, a process known as acute extension of the shoulder may demonstrate sudden onset of substantial weakness in flexion, abduction, and external rotation. As persons age, the rotator cuff becomes increasingly susceptible to tearing with less severe amounts of applied force. Thus, although a major force is required to tear the usual rotator cuff of a 40-year-old person, a relatively trivial force may result in tear of the rotator cuff of the average 60-year-old individual. This is analogous to the predisposition of older women to femoral neck fractures. Although differences of the acromial shape, abnormalities of the acromial-clavicular joint, and other factors may also affect the susceptibility of the rotator cuff to fiber failure, age-related deterioration and loading of the rotator cuff seem to be the dominant factors in determining the failure patterns of the cuff tendons. In a retrospective study of siblings of patients with rotator cuff

Chapter 22  ■  The Rotator Cuff   879

tears, the relative risk of developing symptomatic fullthickness tears in the siblings compared to controls was 4.65. The authors concluded that genetic factors may also play a role in the development of tendon tears in shoulders.18 Symptoms of rotator cuff fiber failure in the acute phase usually include pain at rest and on motion. Later, subacromial crepitance occurs when the arm is rotated in the partially flexed position, and finally, arm weakness occurs. When the rotator cuff fails, shoulder instability can result, and so-called impingement may then manifest. The humeral head is no longer stabilized and may impinge on the tissues between the head and the acromion (acromial process) or between the head and the posterior glenoid in cases of subacromial impingement, the process will lead to osteosclerosis and remodeling of the acromion, and it may result in a traction spur along the coracoacromial ligament.19

TECHNICAL CONSIDERATIONS Mechanical sector scanners with frequencies between 5 and 10 MHz used in the early 1980s provided adequate detail in detecting full-thickness tears.20 The utility of these transducers was limited by several factors: nearfield artifact, narrow superficial image field, and tendon anisotropy. This last artifact is caused by the anisotropic structure of tendons and still affects the scanning of tendons significantly. Parallelism of collagenous structures within the cuff results in peculiar imaging characteristics; the echogenicity of the tendon depends on the angle of the transducer relative to the tendon during tendon interrogation. Oblique insonation of the tissues will result in heterogeneous appearance of the tendons. With optimal perpendicular technique, the center of the image will appear hyperechoic, whereas the side lobes will often be hypoechoic if one scans over the round surface of the proximal humerus. This hypoechogenicity can be mistaken for pathology by the inexperienced ultrasound reader. State-of-the-art imaging of the cuff should be done with a high-resolution linear array transducer with a broad-bandwidth frequency capability, typically 5 to 13 MHz.20 These transducers demonstrate marked improvement in near resolution compared with the older devices. The broad superficial field of view is helpful to improve the near-field image. Tissue harmonics has been shown to improve tendon surface visibility over conventional ultrasound.21 In recent years the ultrasound machines have changed from heavy, space occupying ultrasound equipment to lightweight, laptop ultrasound systems with ergonomic ultrasound probes. Parallel with this trend, the high-expense equipment can often be replaced with more affordable pieces that are marketed for focused use in clinics,20,22 in the operating room, and at the point of injury.23,24 Several manufacturers now

make ultrasound units that weigh less than 5 kg (11 1b); prices of equipment applicable for use in musculoskeletal ultrasound have decreased by more than 80% compared with prices in the 1990s.20

ANATOMY AND SONOGRAPHIC TECHNIQUE Understanding the complex three-dimensional (3-D) rotator cuff anatomy during sonography is crucial to successful rotator cuff sonography. Bone often limits the examination done by the inexperienced operator. For those starting in shoulder ultrasound, but who have experience in MRI arthrography, we recommend performing a quick ultrasound examination before and after each arthrogram. This allows examiners to test their diagnostic abilities instantaneously. Those who have no experience with arthrography can scan in the operating room or anatomy laboratory. Surgical exploration or dissection may teach the most valuable lessons. Those initial steps are necessary to improve knowledge of the anatomy, which is essential in mastering the technique and accelerating the learning curve. Some investigators have been combining arthrographic technique with the sonographic examination, called arthrosonography, which may be more sensitive in assessing synovial proliferation and estimating the size of rotator cuff tears.25 Future applications may also include the diagnosis of labral abnormalities.26-29 As with MRI, ultrasound’s display of anatomy improves when enhanced by injection of intra-articular fluid. Saline used as a contrast agent in arthrosonography is much less expensive than gadolinium, the contrast agent universally used for MR arthrography. The bony landmarks guide the shoulder ultrasound examination (Fig. 22-1). The fingers of the examiner can palpate the acromion, the scapular spine, the coracoid, and the acromioclavicular joint. Transducer orientation relative to those landmarks will be essential in making corrections to the technique in viewing complex shoulder pathology. External bony landmarks are important in shoulder imaging when scanning a patient with significant pathology and loss of normal soft tissue landmarks. The patient is scanned while seated on a rotating stool without armrests. The examiner sits comfortably on a stool adjusted so that the examiner rises above the shoulder level of the patient. Both shoulders, starting with the less symptomatic one, should be examined if the examiner is a beginner. The following technique is used at our institution.8 Transverse images through the long biceps are obtained with the arm and forearm on the patient’s thigh, the palm supinated (Fig. 22-2). The bicipital groove serves as the anatomic landmark to differentiate the subscapularis tendon from the supraspinatus tendon.

880   PART III  ■  Small Parts, Carotid Artery, and Peripheral Vessel Sonography

c

d

e

a

b

c

d

f

A

d

s L

g s S

cb

d

g

D

e

LS

d

f

sc

i

t

cb

tb T

B

a

b d

s

c i t

L

S

Shoulder joint in: Adduction Hyperextention Medial rotation

C FIGURE 22-1.  Right shoulder joint with surface anatomy and underlying musculoskeletal structures. A, Neutral position of the shoulder: frontal view (left image), side view (middle image), and posterior view (right image) of superficial anatomy. The bony prominences formed by the clavicle (a), acromion (b), and coracoid process (c) that limit the acoustic window for shoulder ultrasound are visible subcutaneously. Other visible contours represent the greater tuberosity (d) and spine of the scapula (e), which ends at the medial flat surface (f ), over which slides the aponeurosis of the trapezius muscle. B, Anatomy in neutral position. Frontal view (left image) shows coracoacromial ligament (g), supraspinatus tendon attachment (s), and biceps brachii, with the short head (S) originating from the coracoid medially and the long head (L) extending into the joint deep to the coracoacromial ligament. The coracobrachial tendon (cb) is another tendon that originates from the coracoid. Side view (middle image) shows how the subscapularis (sc) is a separate tendon at the anterior aspect of the shoulder. This tendon is divided from the supraspinatus (S) by the long head of biceps tendon and by the rotator cuff interval. Posterior view (right image) with trapezius (t), infraspinatus (i), teres minor (t), triceps brachii (tb), and teres major (T). The deltoid (D) has been cut, and its edge is seen around the acromion. On all anatomic drawings that follow, the deltoid has been removed. On ultrasound, we look through the deltoid to see the rotator cuff. C, By using adduction, hyperextension and internal (medial) rotation, one can free more supraspinatus tendon for visualization. The most vulnerable zone of the supraspinatus (S) is anterior to the acromion and lateral to the intracapsular long (L) biceps tendon. The zone of the rotator cuff where most tears occur (critical zone) can be found in the trapezoidal space between the bony prominences of the lateral clavicle (a), anterior acromion (b), coracoid (c), and anterior greater tuberosity. The subscapularis tendon in this position hides under the coracoid and medial to the short head of biceps (S).

Chapter 22  ■  The Rotator Cuff   881

Sc

D

B

S Sc

D B LH

C

B

A

D

FIGURE 22-2.  Short-axis or transverse scan of the long biceps tendon. A, View of the long head (LH) and the short head of biceps tendon. Transducer position is indicated by a transparent symbol. For this scan, the patient rests the dorsum of the hand comfortably on the thigh with elbow flexed. S, Supraspinatus tendon; Sc, subscapularis tendon. B, Cross section through the biceps groove at the level of the subscapularis (Sc). The deltoid (D) is hoof shaped and covers the two biceps tendons and the subscapularis tendon in the front of the shoulder. This is the anatomic view of the shoulder in neutral position and with the transducer placed transversely over the bicipital groove. The cross sections of the long and short heads of biceps are seen close together. C, Deep to the deltoid (D), the proximal long biceps (B) rests in the bicipital groove (arrows). The transverse ligament (black arrow) represents the lateral extension of the subscapularis and covers the long biceps. A small segment of the subscapularis is noted between the bone surface of the humerus and the short head of the biceps at the right edge of the image. D, Transverse scan over the lower part of the bicipital groove in a patient with rotator cuff disease. The long biceps tendon (B) appears enveloped in a distended hypoechoic sheath (arrows). The hypoechogenicity of the tendon sheath may represent fluid, synovial hypertrophy, or both. It is important to scan the lowest recess of the biceps synovial sheath. In a patient who sits for the examination, fluid will precipitate to the most dependent portion of the synovium.

The groove is concave; bright echoes reflect off the bony surface of the humerus. The tendon of the long head of the biceps is visualized as a hyperechoic oval structure within the bicipital groove on the transverse images. The tendon courses through the rotator cuff interval and divides the subscapularis from the supraspinatus tendon. Scanning should begin with the proximal long biceps tendon above the biceps tendon groove. The intracapsular biceps shows more obliquely in the shoulder capsule. The capsular biceps is located in a space typically referred to as the rotator cuff interval. The space varies between 1 and 3 cm in width.30 In this interval between the superior subscapular and the anterior supraspinatus, a sling of connective tissue surrounds the proximal long biceps tendon. Deep to the biceps, inserting on the

bicipital groove, the sling consists of fibers of the superior glenohumeral ligament. Superficial to the biceps are fibers of the coracohumeral ligament,31 which courses from the coracoid medially, covers the biceps in the rotator cuff interval, and attaches on the humerus laterally. At its lateral insertion, the coracohumeral ligament forks around the anterior supraspinatus. The deep layer appears more distinct and has been called the rotator cuff cable,32 a structure at the articular margin of the supraspinatus. The superficial coracohumeral ligament is thinner and less distinctly visible. The sonographers may occasionally recognize the rotator cuff cable because of its unique anisotropic characteristics oriented perpendicular to the longitudinal fibers of the critical zone of the supraspinatus. After scanning the biceps in the

882   PART III  ■  Small Parts, Carotid Artery, and Peripheral Vessel Sonography

capsule, the long biceps is followed throughout its course in the bicipital groove; the scan should extend as far down as the musculotendinous junction. This allows detection of the smallest fluid collections in the medial triangular recess at the distal end of the tendon sheath.33 Such small biceps sheath collections are a very sensitive indicator of joint fluid. A 90-degree rotation of the transducer into a longitudinal view will ascertain the intactness of the biceps tendon.34 The transducer must be carefully aligned along the biceps groove (Fig. 22-3). Gentle pressure on the distal aspect of the transducer is necessary to align the transducer parallel to the tendon to avoid artifact due to anisotropy (Fig. 22-4). The transducer position is then returned to the transverse plane and moved proximally along the humerus to visualize the subscapularis tendon, which appears as a band of medium-level echoes deep to the subdeltoid fat and bursa. The subscapularis tendon is viewed parallel to its axis (Fig. 22-5) for its long-axis view; scanning during passive and external rotation may be helpful in assessing the integrity of the subscapularis tendon, which may be disrupted in patients with chronic anterior shoulder dislocation. External rotation is also necessary to diagnose subluxation of the long biceps tendon, especially if only present intermittently.35 The short-axis view of the subscapularis is best evaluated over the ledge of the lesser tuberosity as close as possible to the bicipital groove (Fig. 22-6).

The normal subdeltoid bursa is recognized as a thin, hypoechoic layer between the deltoid muscle on one side and the rotator cuff tendons and biceps tendon on the deep side. Hyperechoic peribursal fat surrounds the outer aspect of the synovial layer.36 The supraspinatus tendon is scanned perpendicular to its axis (transversely) by moving the transducer laterally posteriorly. The sonographic window is very narrow, and careful transducer positioning is essential (Fig. 22-7). The supraspinatus tendon is visualized as a band of medium-level echoes deep to the subdeltoid bursa and superficial to the bright echoes originating from the bone surface of the greater tuberosity. The rest of the examination is done with the arm adducted and hyperextended and the shoulder in moderate internal rotation5,7,37 (Fig. 22-8). This position can best be explained to patients by asking them to reach to the opposite back pocket. This placement of the hand is often alternated with the hand position over the back pocket on the side of the affected shoulder with maximum elbow adduction. The latter position puts more tension on the tendon. It may make a lesion of the cuff stand out more clearly; however, the position may also mislead the examiner to overestimate the size of tears.38 Both longitudinal sections along the course of the supraspinatus tendon and images transverse to the tendon insertion and perpendicular to the humeral head are obtained. Correct orientation is achieved when an imaging plane

LH

B LH

B

A

C

FIGURE 22-3.  Long-axis or longitudinal scan of the long biceps tendon. A, In the neutral position, the long biceps is found in the bicipital groove, about midline over the anterior humerus. Transducer position is indicated by the transparent symbol. B, Longitudinal cross section through the biceps. The deltoid is also hoof shaped in the sagittal plane. This is the anatomic view of the shoulder in neutral position and with the transducer placed longitudinally over the bicipital groove. The long head of the biceps (LH) appears as a tubular structure. C, Longitudinal scan shows the biceps tendon (B) with distinct fibrillar architecture deep to the deltoid. Note the distinct longitudinal linear reflections in the tubular structure of a normal long biceps. When the layered structure is absent, the clinician should consider the possibility of scar tissue replacing a torn tendon.

Chapter 22  ■  The Rotator Cuff   883

FIGURE 22-4.  Composite of images through proximal long biceps showing tendon anisotropy. The predominant longitudinal orientation of the collagen in tendons such as the biceps make them strong anisotropic reflectors. Top images, Transverse imaging approach; bottom images, sagittal scanning through the middle of the bicipital groove. The column on the left shows correct scanning technique. Normal tendons appear hyperechoic only when scanned perpendicularly. The column on the right demonstrates how tendons appear hypoechoic as the angle of the transducer diverges from 90 degrees. Visualization of this transition of tendon echogenicity from hyperechoic to hypoechoic may be used at times to improve tissue contrast; it is also a useful technique to distinguish tendon from scar.

shows crisp bone surface definition and sharp outline of the cartilage of the humeral head. During longitudinal scanning, the transducer overlays the acromion medially and the lateral aspect of the greater tuberosity laterally (Fig. 22-8, B and C). The transducer sweeps around the humeral head circumferentially, and the transducer should be held perpendicular to the humeral head surface at all times. This sweeping motion through the supraspinatus tendon starts anteriorly next to the long biceps tendon. We cover an area of approximately 2.5 cm lateral to the long biceps tendon. Infraspinatus tendon is scanned beyond this point. The musculotendinous junction shows as hypoechoic muscle surrounding hyperechoic infraspinatus tendon. The transverse scan starts just lateral to the acromion and translates downward over the supraspinatus tendon and the greater tuberosity. The critical zone is that portion of the tendon that begins approximately 1 cm posterolateral to the biceps tendon. Failure to adequately visualize this area may cause a false-negative result.5

Scanning of the supraspinatus tendon is followed by the visualization of the infraspinatus and teres minor tendons by moving the transducer posteriorly and in the plane parallel to the scapular spine. The infraspinatus tendon appears as a beak-shaped soft tissue structure as it attaches to the posterior aspect of the greater tuberosity6 (Fig. 22-9). Internal and external shoulder rotation may be helpful in the examination of the infraspinatus tendon. This maneuver relaxes and contracts the infraspinatus tendon in alternating fashion. At this level, a portion of the posterior glenoid labrum is seen as a hyperechoic triangular structure. The fluid of the infraspinatus recess surrounds the labrum. Optimal image contrast for detection of intra-articular fluid will be obtained by bringing the arm in external rotation (Fig. 22-10). In this position the normal labrum will be covered by infraspinatus tendon. Both structures appear hyperechoic and become almost indistinguishable in a joint without effusion. In contrast, hypoechoic fluid or synovium may considerably separate these tissues in the

884   PART III  ■  Small Parts, Carotid Artery, and Peripheral Vessel Sonography

Sc D

Sc

B D Sub

B

A

Cor

C Arm in neutral position

Sc

L

D

Sc

FIGURE 22-5.  Long-axis view of subscapularis tendon, or transverse scan through anterior shoulder. A, Transducer position for the examination of the long axis of the subscapularis (Sc) is indicated by the transparent transducer. B, Cross section of the subscapularis (Sc) and the two tendons of the proximal biceps. The subscapularis tendon can be seen over the humerus and between the two biceps tendons. In neutral position or in internal (medial) rotation, there is little separation of the two biceps tendons. In extreme rotation, the long head may actually pass under the short head. C, Long-axis sonogram visualized with external rotation shows the subscapularis tendon (Sub) parallel to its axis, viewed as a band of medium-level echoes deep to the deltoid muscle (D); B, biceps tendon; Cor, coracoid. D, Dual images illustrate the use of external rotation in bringing out the subscapularis from under the coracoid and short head of biceps (S). In the neutral position, long head (L) of the biceps will show over the middle of the proximal anterior humerus. In external rotation, long head of the biceps separates from the short head. The subscapularis tendon (Sub) shows in its full length over the anterior humeral head; D, deltoid. Anatomic diagrams on the left show the long biceps position in frontal view. Arrows indicate what happens to the cross-sectional anatomy with internal rotation (top) and external rotation (bottom).

S

L

S

S

L

Arm in lateral rotation

Sc L L

S

S D

Sc

Sub L

D

S

D L

Sc L

Sc

B D

D SUB

C

A

FIGURE 22-6.  Short-axis view of subscapularis, or sagittal scan through anterior shoulder. A, The transparent transducer indicates the placement of the transducer. B, The subscapularis tendon (Sc) covers the humeral head and the humeral head cartilage of the anterior shoulder. The tendon is visualized through the deltoid (D). The space superior to the subscapularis, the rotator cuff interval, is a space with loose mesenchymal tissue surrounding the intracapsular biceps (L). C, Deep to the deltoid (D), the subscapularis (SUB) tapers in thickness from its superior to inferior border.

D S. Bursa

S

S

B D

A

D

SUP

SUP

HH

HH

C

FIGURE 22-7.  Short-axis scan of supraspinatus tendon. A, With the arm in extension and internal rotation, the transducer is placed between the anterior acromion and the coracoid. The transducer is swept from the edge of the acromion down to the level of the lateral greater tuberosity. B, This section in the plane of the transducer shows how the supraspinatus is covered most immediately by the subdeltoid bursa (S. Bursa) deep to the deltoid (D). C, The supraspinatus tendon (SUP) shows as a band of medium-level echoes deep to the subdeltoid bursa (arrows) and draped over the cartilage of the humeral head (HH). The swollen symptomatic tendon shows on the left side of the split-screen image; the patient’s normal shoulder shows on the right. D, Deltoid muscle. Compare the thickness of the hypoechoic bursa with the thickness of the hyaline cartilage covering the humeral head. A normal bursa does not exceed the thickness of normal cartilage.

886   PART III  ■  Small Parts, Carotid Artery, and Peripheral Vessel Sonography

Deltoid S

S C

D

C

S D

S

Acr

Hu

D

B Subdeltoid bursa

D S

IS

A

E

CI

SUP

L

IS HH

Sc

B SUB

F

FIGURE 22-8.  Long-axis scan of supraspinatus tendon. A, With the arm in extension and internal rotation, the transducer is placed perpendicular to the curvature of the acromial process for the view along the bulk of the largest number of fibers of the supraspinatus. We refer to this view as the long-axis view. B, This section demonstrates the challenge posed by the acromioclavicular joint and the lateral acromion that cover part of the supraspinatus (S). The positioning of the patient (as seen in A, inset) frees the critical zone and moves it into the acoustic window deep to the acromioclavicular ligament. C, Longitudinal image through a left supraspinatus. More tendon will stretch beyond the lateral and anterior aspect of the acromion than in the neutral position. Echogenicity changes within the rotator cuff are related to tendon anisotropy. The propagation of ultrasound through supraspinatus tendon (S) appears uneven. This feature stands out more clearly in one sublayer of the supraspinatus tendon (arrows). The fibers in this layer have a longitudinal orientation along the long axis of the tendon; C, hyaline cartilage over the humeral head. D, Panoramic overview through a right supraspinatus demonstrates the anatomic relationship of the longitudinal supraspinatus tendon (S) with the acromion (Acr) and acromioclavicular joint (open arrow). This image can best be viewed side by side with B. Deltoid (D) originates from the acromion; Cl, lateral clavicle; Hu, proximal humerus. The shape of the tendon has often been compared to a parrot’s beak. Hypoechoic tissue covers the tendon on either side. Hypoechogenicity between tendon and bone represents hyaline cartilage (white arrow), and the thin, hypoechoic layer between tendon and deltoid corresponds to subdeltoid bursa (black arrow). E, Sagittal section shows relationship of supraspinatus joining the infraspinatus, which is located more posteriorly. Also noted is the separation of the anterior supraspinatus from the subscapularis (Sc) by the intracapsular long biceps (L). F, Panoramic overview of the transverse anatomy of supraspinatus (SUP) relative to the subscapularis (SUB) and the intracapsular biceps (B) in the front and the infraspinatus (IS) in the back. Again, the tendon appears sandwiched between two hypoechoic layers. Note that the normal subdeltoid bursa (black arrow) remains slightly thinner than the hyaline cartilage (white arrows) over the humeral head (HH). The supraspinatus and infraspinatus form a conjoined tendon, whereas the biceps tendon separates supraspinatus and subscapularis. The image has been made by making a circular sweep over the tendons in a movement following the hoof shape of the deltoid in the sagittal plane, as illustrated in E.

joint with arthritis. The hypoechoic articular cartilage of the humeral head, which shows lateral to the labrum, contrasts significantly with the hyperechogenicity of the fibrocartilage. Scanning is extended medially to encompass the spinoglenoid notch and the suprascapular vessels and nerve. Visualization of the notch may be improved by bringing the transducer in the transverse plane, but with the medial end of the transducer slightly more cephalad than the lateral end. Using the external-internal rotation dynamic during visualization of the neurovascular bundle that wraps around the spinoglenoid notch will show distention of the suprascapular vein during external rotation. The transversely oriented transducer is moved distally, and the teres minor is then visualized.

The teres minor tendon is a trapezoidal structure39 (Fig. 22-11), differentiated from the infraspinatus tendon by its broader and more muscular attachment. Tears of the teres minor are rare. In cases of quadrilateral space syndrome with entrapment of the axillary nerve, the teres minor can be unilaterally smaller and appear hyperechoic.40 Small joint effusions will also be imaged in this location.41 Demonstration of this effusion helps distinguish articular processes, such as rheumatoid arthritis and septic arthritis, which will cause effusion. In rotator cuff disease, it is rare to find fluid in this location. Coronal images through the acromioclavicular joints are obtained at the end of the examination. Right-left comparison can show degenerative or traumatic pathol-

Chapter 22  ■  The Rotator Cuff   887

D

Trapezius

IS D

IS

B

D Hu

IS

Scapula

A

C

FIGURE 22-9.  Long-axis view of the infraspinatus. A, With the arm at the patient’s side, the operator can scan the infraspinatus (IS) from its origin on the scapula to its insertion on the posterior greater tuberosity; D, deltoid. The transducer is placed over the posterior joint (on the left of the transducer) and oriented toward the tuberosity insertion. B, Oblique anatomic section along the infraspinatus (IS) long axis; D, deltoid. C, Panoramic view along the plane shown in B. The infraspinatus (IS) thins out toward its humeral insertion (Hu) and appears sandwiched between the deltoid (D) and the humerus. Arrow indicates the location of the joint.

ogy that can mimic or cause impingement-like symptoms. The superior glenoid labrum can be shown with the transducer aligned posterior to the acromioclavicular joint and oriented perpendicular to the superior glenoid. A curved, linear array transducer will be necessary if diagnosis of superior labral detachment (SLAP lesions) is sought.

THE NORMAL CUFF The Adolescent Cuff The rotator cuff tendons and the intracapsular biceps are hyperechoic relative to the deltoid muscle bellies (Fig. 22-12). The cuff tendons are enveloped in a thin synovial layer that is normally thinner than 1.5 mm and appears hypoechoic relative to the tendons. The thickness of this bursal layer does not change. The subacromial-subdeltoid bursa is as thick over the long biceps tendon as it is over the subscapularis, supraspinatus, and infraspinatus tendons. A correctly performed examination will show a neatly defined bursa that shows as a hypoechoic stripe thinner than the thickness of the hypoechoic hyaline cartilage over the humeral head. This extra-articular bursa is a virtual space, because it contains lubricant synovial fluid; this fluid cannot be distinguished on a routine shoulder ultrasound study. The

bursa is hoof shaped in cross section, and it often extends from the coracoid anteriorly around the lateral shoulder and posteriorly past the glenoid. If the subdeltoid bursa extends that far anteriorly, it is in direct continuity with the coracobrachial bursa. The pleural space and the bursal synovial space have a number of similarities, including the virtual space (which can become distended in effusions), the thin lubricating layer of fluid in their lumen, and the extensive network of capillary vessels and lymph vessels in their walls. Those vessels are not visible with color flow Doppler sonogram in patients with normal rotator cuff anatomy, but distended vessels have been shown in the power and color flow Doppler sonograms of patients with inflamed cuffs.42 The boundary between the bursa and the deltoid muscle consists of the so-called peribursal fat. This layer appears hyperechoic, and its thickness is remarkably uniform; body habitus seems to have little influence on the thickness of this fat layer. Rotator cuff pathology is rare in young patients, although bursal and labral pathology can occur. Some of these conditions can mimic tendon tears. It is important to know that the adolescent cuff consists of more muscle than the aging cuff. The relative length of tendon to muscle increases with age.43 Hypoechoic areas in the cuff in patients under age 20 may simply represent muscle, and the finding should not easily be attributed to a tear. Meticulous right-left comparison of the thickness of the

888   PART III  ■  Small Parts, Carotid Artery, and Peripheral Vessel Sonography INTERNAL ARM ROTATION

D L

IS

HH

M

Glenoid T EXTERNAL ARM ROTATION

IS

L T

Glenoid

HH

RT

EXTERNAL ARM ROTATION WITH INTRAARTICULAR EFFUSION

L E T

FIGURE 22-10.  Relationship of infraspinatus tendon and muscle relative to glenoid labrum. With the arm in medial (internal) and lateral (external) rotation. Top diagram, Fibrous glenoid labrum (L) appears as a hyperechoic triangle deep to the infraspinatus muscle (M). Look at the corresponding ultrasound image on the left. With the arm in internal rotation, the contrast between the hypoechoic muscle (IS) and the labrum (arrows) is accentuated. Middle diagram, When the arm is in external rotation, the hyperechoic tendon (T) is in direct proximity to the hyperechoic labrum. Look at the ultrasound inset on the left. The separating interface, which is barely perceptible (black arrow), represents the actual location of the synovium. Bottom diagram, External rotation is the preferred position to detect effusion. If there is effusion (E), hypoechoic fluid distends the synovium between the posterior labrum and the deep surface of the infraspinatus. HH, Humeral head.

subscapular tendons in adolescents may demonstrate tears of the anterior rotator cuff resulting from athletic injuries. In our experience, the subscapular insertion appears the weaker link of the rotator cuff in the growing shoulder. Ultrasound has proved its usefulness in detecting subscapularis tendon tears.44

Age-Related Changes The rotator cuff in individuals under age 30 years is watertight. Arthrographic studies show that no commu-

nication should exist between the glenohumeral joint and the subacromial-subdeltoid bursa.25 Postmortem and cadaver studies have shown a high prevalence of rotator cuff tears in aging shoulders. Keyes45 examined 73 unselected cadavers and found full-thickness tears of the supraspinatus in 13.4% of shoulders. Full-thickness tears were not recorded for those younger than age 50 years; the prevalence over age 50 was 31%. Wilson and Duff46 examined an unselected series of 74 bodies at postmortem and 34 dissecting-room cadavers over age 30 years. They found full-thickness tears of the supra-

Chapter 22  ■  The Rotator Cuff   889

D

Trapezius

TM IS

D

B TM

D

IS

D

D

TM

A

C

FIGURE 22-11.  Long-axis view of the teres minor. A, The probe placed parallel to the spine of the scapula and just proximal to the prominence caused by the muscle belly of the teres major (look at the surface landmarks in Fig. 22-1) can show the short tendinous insertion of the teres minor (TM); IS, infraspinatus; D, deltoid, severed proximally. B, Obliquely along the teres minor insertion, diagram shows the “muscular” (fleshy) insertion of the teres minor tendon; TM, teres minor; IS, infraspinatus. C, Teres minor (TM) is seen as a trapezoidal structure deep to the inferior deltoid (D); humerus (arrows). The insertion appears hypoechoic, as opposed to other rotator cuff tendon insertions that have more fibrinous and therefore more hyperechoic insertions.

Deltoid

Rotator cuff B

Biceps C

A

B

FIGURE 22-12.  Short-axis view of supraspinatus and rotator cuff interval. A, In young and healthy individuals, the rotator cuff contrasts significantly in echogenicity with the deltoid. The cuff tendons appear hyperechoic relative to the deltoid. C, Coracoid process. B, In the rotator cuff interval, a hyperechoic “sling” surrounds the biceps. Fibers of the superior glenohumeral ligament (white arrow) are seen deep to the intracapsular biceps (B), and fibers of the coracohumeral ligament (black arrows) course superficially.

spinatus tendon in 11% and partial-thickness tears in 10% of the shoulders. Fukuda et al.47 reported a 7% prevalence of complete tears and a 13% prevalence of incomplete tears in a study of cadavers that included no details on age. With such high percentages of rotator cuff tears in cadaver studies, how many of these tears would have been asymptomatic? A study we conducted showed that ultrasound can detect asymptomatic tears. Ninety volunteer subjects (47 women and 43 men) in a population who had never sought medical attention

for shoulder disease underwent shoulder sonography; 77% (69 of 90) were white, 13% (12 of 90) were black, 9% (8 of 90) were Asian, and 1% (1 of 90) was Hispanic. Eighteen subjects were between ages 30 and 39 years; 18 were 40 to 49 years old; 18 were 50 to 59; 13 were 60 to 69; 13 were 70 to 79; and 10 were 80 to 99 years old. The proportion of women to men was almost equal for each decade. No statistically significant differences were found in the prevalence of rotator cuff lesions in each gender for

890   PART III  ■  Small Parts, Carotid Artery, and Peripheral Vessel Sonography ROTATOR-CUFF CHANGES IN ASYMPTOMATIC ADULTS DOMINANT VERSUS NON-DOMINANT

ROTATOR-CUFF CHANGES IN ASYMPTOMATIC ADULTS IMPINGEMENT GRADES

100

100 Dominant arm Non-dominant

80

Percentage

Percentage

80

60

40

Years

60

40

20

20

0 Decade

Stage 1 Stage 2 Stage 3

4

5

6

7

8

9 and 10

30–39

40–49

50–59

60–69

70–79

80–99

FIGURE 22-13.  Asymptomatic rotator cuff tears. Percentage of shoulders with rotator cuff tears in asymptomatic adults in different age groups. Chart shows comparison between dominant and nondominant arms.

either the dominant or the nondominant arm (Fig. 22-13). We found no statistically significant differences in the incidence of rotator lesions related to gender or reported level of exertional activities. However, the prevalence of rotator cuff tears in dominant and nondominant arms showed a linear increase after the fifth decade of life. This difference was statistically significant among patients in the third, fourth, and fifth decades and older.13 The cumulative percentage of partial-thickness and full-thickness tears was approximately 33% between ages 50 and 59 years, 55% between 60 and 69 years, 70% between 70 and 79 years, and as high as 78% above age 80 (Fig. 22-14). A total of 25 full-thickness and 15 partial-thickness tears were found. Sixteen individuals or 64% of the people with tears had bilateral rotator cuff tears. The youngest subject with a partial-thickness tear was 35 years old. The youngest subject with a fullthickness tear was 54 years old. The age range for partialthickness tears was 35 to 80. The age range of full-thickness tears was 54 to 92. The average age in the partial-thickness group was 56 years. The average age in the full-thickness group was 63 years. In 19 cases (46%) the rotator cuff tears had associated intrasynovial fluid. In 15 of these patients the fluid was located in the biceps tendon sheath, and in the remaining four cases it was located in the subacromial-subdeltoid bursa. There were two individuals with tears and fluid in the biceps tendon sheath and in the bursa simultaneously. The infraspinatus recess appeared normal in all our patients. Eleven effusions were noted in the long biceps tendon sheath in subjects who did not have tears of the cuff. There was never excess fluid in the subacromial-subdeltoid bursa in the absence of rotator cuff tear.

0 Decade Years

4

5

6

7

8

9 and 10

30–39

40–49

50–59

60–69

70–79

80–99

FIGURE 22-14.  Prevalence of stage 1 to stage 3 impingement for dominant arm in different age groups. Abnormalities in the subacromial space were staged sonographically as follows: stage 1 if bursal thickness 1.5 to 2 mm; stage 2 if bursal thickness over 2 mm; stage 3 if partialthickness or full-thickness rotator cuff tear.

D

b gt ss

+

FIGURE 22-15.  Bone change in asymptomatic fullthickness rotator cuff tear. Longitudinal scan through the supraspinatus tendon (ss) shows retraction of tissue (calipers). The bone surface of the uncovered greater tuberosity (gt) is irregular. The subdeltoid bursa (b) is filled with fluid. D, Deltoid muscle.

Shallow erosion or irregularity of the bone surface under the tear was noted in 90% of tears; bone changes were present in all but four partial-thickness tears. Greater tuberosity irregularity was noted in 37 shoulders or in 21% of shoulders in this study. Twelve shoulders showed irregular greater tuberosities and no rotator cuff tear. A statistically significant correlation between asymptomatic rotator cuff tears and irregularity of the greater tuberosity was found (Fig. 22-15). Twenty of the full-thickness tears were considered large and involved more than one tendon. Three tears

Chapter 22  ■  The Rotator Cuff   891

were massive and greater than 4 cm in diameter, and three tears were small and less than 2 cm in width when measured over the base of the greater tuberosity. Ten partial-thickness tears were mixed echogenicity, and five were hypoechoic. Nine mixed-echogenicity lesions and two hypoechoic tears exhibited bone change in the greater tuberosity. Our results indicate that the finding of a rotator cuff abnormality or an effusion in the biceps tendon sheath can be compatible with normal and pain-free mobility of the shoulder. Rotator cuff findings should be interpreted with care in patients over age 50. A rotator cuff tear is not necessarily the cause of the pain in an aging shoulder and can be an incidental finding. Degenerative rotator cuff changes may be regarded as a natural correlate of aging, with a statistically significant linear increase after the fifth decade of life. On the one hand, clinical judgment must be used to distinguish asymptomatic from symptomatic rotator cuff tears. On the other hand, finding a rotator cuff tear should not stop the clinician from searching for other causes of shoulder pain. Our shoulder ultrasound reading is done in conjunction with the reading of the initial shoulder radiographic evaluation. We have found missed primary or secondary neoplasms of bone, myeloma, and Pancoast tumors using this careful approach. Limited and painful shoulder elevation can result from a number of diseases, of which rotator cuff disease is the most common. Simultaneous occurrence of a full-thickness tear with a tumor in or around the shoulder is not rare in our experience. Yamaguchi et al.48 studied the contralateral asymptomatic arm of patients with symptomatic tears in one arm. In a follow-up averaging 2.8 years, 51% of asymptomatic shoulders became symptomatic. In 23 patients the tears were reevaluated with ultrasound. None of the tears had healed, and none had become smaller. In 9 of the 23 patients the tears had increased in size. In patients with tears that were first asymptomatic and then symptomatic, the ability of performing daily activities decreased significantly. In a study of middle-aged tennis players, Brasseur et al.49 showed that tears of the rotator cuff were more than twice as common in the dominant arm. The tears discovered by ultrasound had been symptomatic at one time or other in 90% of the players. However, there was no relationship between the presence of a tear or calcification at the time of the study and the presence or absence of pain.4

PREOPERATIVE APPEARANCES Criteria of Rotator Cuff Tears Rotator cuff ultrasound has become more popular partly because the imaging of the rotator cuff has been per-

fected with a high degree of sophistication. In addition, patients and clinicians have contributed to the recent surge in interest in shoulder ultrasound. Patients who have undergone both ultrasound and MRI of the shoulder prefer ultrasound over MRI.50,51 Clinicians who have thorough knowledge of shoulder anatomy and pathology now have access to compact ultrasound technology. These physicians see the advantages with in-office ultrasound; the technique is low cost and provides the opportunity for patient education during the visit.20,22 With respect to the reproducibility of the study of the rotator cuff, several radiologists have tested agreement between experienced radiologist readers and have found good to excellent interobserver agreement for fullthickness rotator cuff tear evaluation (kappa values between 0.6 and 0.81). In cases of partial-thickness tears and for intratendinous changes, the interobserver variability was higher.52-54 O’Connor et al.54 noted poor agreement between an experienced operator and a less experienced operator with only 6 months of training in shoulder ultrasound. This study concluded that rigorous training with measurement of competency will be required if the medical community wants to keep this technique at its current degree of diagnostic credibility. Previously published sonographic criteria for rotator cuff pathology can be categorized into four groups: nonvisualization of the cuff, localized absence or focal nonvisualization, discontinuity, and focal abnormal echogenicity.55

Nonvisualization of the Cuff Direct contact of the humeral head with the acromion is an indication of massive cuff tear. In this situation the ultrasound image shows deltoid muscle directly on top of the humeral head (Fig. 22-16). In some cases, thickened bursa and fat will be noted between the deltoid muscle and the surface of the humeral head. This tissue layer is more hypoechoic and patchy in texture. The thickness of this layer will depend on the location of the tear, but generally it will be thinner and more irregular than the normal cuff layer. Some bursae have been noted to be up to 5 mm thick. This synovial layer has been mistaken for normal cuff by the inexperienced sonographer. With massive tears, exceeding 4 cm, the humeral head may ascend through the defect because of pulling of the deltoid muscle. The supraspinatus tendon is retracted under the acromion, and as a rule, surgical reattachment will be challenging at this stage (Fig. 22-17). The extent of tear should be reported because multiple tendons are often involved. The diagnosis of these tears can be predicted on shoulder radiographs. Some centers use radiographs with comparison views during active shoulder abduction or anteroposterior supine views of the subacromial space to counteract the gravitational pull on the humerus.56 The subacromial space should not be smaller than 5 mm.

892   PART III  ■  Small Parts, Carotid Artery, and Peripheral Vessel Sonography

D

D

H

H

INFRASP

A

B

FIGURE 22-16.  Nonvisualization of cuff. A, Transverse view shows the deltoid muscle (D) in direct contact with the humeral head (H). Hyperechoic layer (curved arrow) of fat lies deep to the deltoid; this layer is interposed between deltoid and humerus. B, Longitudinal view through the expected location of the supraspinatus tendon; supraspinatus is absent. Hyperechoic layer of fat (curved arrow) is noted deep to the deltoid (D). H, Humeral head.

SS

RHT

COMP LFT

FIGURE 22-17.  Irreparable rotator cuff tear. Longitudinal right-left comparison shows a significant discrepancy in the thickness of the soft tissues. The supraspinatus tendon (SS) appears normal in the asymptomatic left shoulder (LFT). The supraspinatus tendon in the right (RHT) shoulder is retracted out of sight. The deltoid and the subdeltoid fascial layer cover the humeral head directly. Arthroscopy showed the torn edge of the supraspinatus tendon withdrawn beyond the glenoid cavity. The rotator cuff defect was deemed irreparable.

Focal Nonvisualization of the Cuff Smaller tears will appear as localized absence of supraspinatus tendon or, in rare cases, local absence of subscapularis or infraspinatus tendon. The most common tear pattern is caused by disease at the tendon-bone junction. The tendon will retract from the bone surface, leaving a bare area of bone (Fig. 22-18). This finding has been reported in the past as the “naked tuberosity” sign.57 The bone surface of the greater tuberosity and anatomic neck of the humerus are irregular in approximately 79% of this type of tears; anatomic study confirmed these bone changes.58 This pathologic process affects not only the surface of the bone, but also the internal structure of the greater tuberosity. The exterior changes consist of pitting of the cortex, erosion of bone,

sclerosis, fragmentation of the tuberosity, and crystal deposition beyond the tidemark.58 The changes of the architecture of the tuberosity manifest as fewer trabeculae and fewer connections between trabeculae. The vast majority of such tears will occur anteriorly in the supraspinatus tendon and in the critical zone. Characteristically, a small amount of tissue will be preserved surrounding the biceps tendon. Ideally, such tears can be confirmed in two perpendicular scan planes. Sometimes this will not be possible because the tear may show full thickness in one plane but may not be identified as such in the orthogonal plane. This phenomenon has been attributed to partial-volume averaging in tears that are smaller than the footprint of the transducers. Small horizontal tears typically appear on longitudinal images but can be missed on transverse images.57 A helpful

Chapter 22  ■  The Rotator Cuff   893

finding is the “infolding” of bursal and peribursal fat tissue into the focal defect. With few exceptions, this infolding is a sign of a full-thickness tear. If the tear is larger, bursal and peribursal tissue will approximate the bone surface (Fig. 22-18). Large bursal surface tears can occasionally show this pattern of infolding.50 Focal nonvisualization should not be confused with segmental thinning of cuff after rotator cuff surgery. This thinning is normal after most tendon-bone reimplantations. In these patients a bony trough is detected as a rounded or V-shaped defect in the humeral contour. The tendon is brought down into this narrow slit. The

RT

S gt

E

FIGURE 22-18.  Horizontal full-thickness tear. Longitudinal image through the supraspinatus tendon (S) shows 2-cm retraction of the torn tendon (distance between calipers). Bursa and peribursal fat (curved arrow) rest directly on the irregular bone surface of the greater tuberosity (gt). E, Humeral epiphysis.

tendon is not repaired onto the tuberosity anatomically with a broad insertion but with a tapered end. It is well known that a number of these reconstructions fail to be watertight even after successful surgery. The rents in the capsule cause additional focal thinning. In a patient with a negative baseline study, re-tears can be identified by visualizing anechoic fluid leaking through a tear.

Discontinuity in the Cuff The term discontinuity has been used for tears located more proximally in the tendon. These tears tend to be of the vertical type and are more often traumatic.57,59 The patient may have a history of prior shoulder dislocation. Discontinuity is observed when the small defects fill with joint fluid or hypoechoic reactive tissue60 (Fig. 22-19). Such defects are often accentuated by placing the arm in extension and internal rotation (Fig. 22-20). Often, a small amount of bursal fluid is also present. The sonographer can use this fluid as a natural contrast medium to show the tear in more detail. Manual compression of the subdeltoid bursa can move the fluid through the tear into the joint. This maneuver will show the tear more clearly. A focally bright interface around a segment of hyaline cartilage and deep to hypoechoic tendon is considered a sign of a full-thickness tear (see Fig. 22-19). This sign has been named the cartilageinterface sign in an earlier report.57

Focal Abnormal Echogenicity Cuff echogenicity may be diffusely or focally abnormal. Diffuse abnormalities of cuff echogenicity have proved to be unreliable sonographic signs for cuff tear, especially

De

De

Su Su

FIGURE 22-19.  Vertical full-thickness tear. Left, Longitudinal, and right, transverse, split-screen images through the supraspinatus tendon (Su) show an anechoic area of discontinuity (large arrows) within the rotator cuff layer. The cartilage of the humeral head is surrounded by a bright interface (small arrows). De, Deltoid muscle.

894   PART III  ■  Small Parts, Carotid Artery, and Peripheral Vessel Sonography

D D

D

D

D SUP SUP

A

SUP

B

FIGURE 22-20.  Discontinuity of the cuff. A, Transverse scans of the supraspinatus tendon (SUP) in neutral position. B, Transverse scans of the supraspinatus tendon with the arm in extension and internal rotation show a small tear filled with fluid (arrowheads). Note the tear is more distinctly visible with the arm in extension. D, Deltoid muscle; arrows, subdeltoid bursa.

A

B

FIGURE 22-21.  Focal abnormal echogenicity. A, Longitudinal supraspinatus tendon view of an articular side partial-thickness tear, the so-called rim rent. A linear hyperechoic lesion in the supraspinatus tendon (open arrow) is surrounded by hypoechoic edema (curved arrows). B, Transverse supraspinatus tendon view of same partial-thickness defect. The same hyperechoic lesion is noted.

when there is no associated bone surface change.61 Focal abnormal echogenicity has been associated with small full-thickness and partial-thickness tears. An area of increased echogenicity might represent a new interface within the tendon at the site of fiber failure, as observed in some partial-thickness tears.8 The small, linear or comma-shaped hyperechoic lesion is often surrounded by edema or fluid and appears as a hypoechoic halo (Fig. 22-21). The partial-thickness tears are similar to the rim rents first observed pathologically by Codman.10 A slightly different type of partial-thickness tear can appear as an anechoic spot on the articular or bursal side of the tendon.8 Careful inspection of the synovial surfaces of the tendon is necessary. Only those focal hypoechoic

defects that violate the surface may be considered tears by the arthroscopist (Fig. 22-22). Intrasubstance lesions are the most common type of partial lesions and account for almost 50% of the defects. We do not call them “tears” because they are not considered tears by the surgeons who cannot observe them by direct tendon inspection. This poses a diagnostic problem similar to that of intrasubstance lesions of the menisci seen on MRI studies. Associated bone or synovial findings may be helpful if the ultrasound findings are equivocal.61 Tissue harmonics makes the intratendinous cleavage or delamination stand out more visibly. Because of this new technology, diagnostic accuracy for partial-thickness tear has recently improved.62

Chapter 22  ■  The Rotator Cuff   895 Right shoulder

B

A

B Right shoulder

C

Associated Findings Subdeltoid Bursal Effusion Visualization of subdeltoid bursal effusion is the most reliable associated finding of rotator cuff tear (Fig. 22-23). It is found in both full-thickness and partialthickness tears. Anechoic fluid differs from hypoechoic edema of the bursal synovium. Edema is a common finding in shoulder impingement but is only rarely associated with a tear. Edema and fluid can be distinguished from each other using the transducer compression test. A synovial recess filled with fluid will be emptied by compression; a recess with synovial edema changes little in shape. Other causes for fluid in the bursa include calcium milk with synovitis and septic bursitis. Hollister et al.63 found that the sonographic appearance of bursal fluid had a specificity of 96% for

FIGURE 22-22.  Focal abnormal echogenicity. A, Transverse supraspinatus tendon view of a bursal-side partialthickness tear. The hypoechoic change violates the bursal surface (arrows). B, Transverse supraspinatus tendon view of hypoechoic change within the substance of the tendon. C, Longitudinal view through the same abnormality as in B. The hypoechoic disruption appears intrasubstance (arrows); intact tendon fibers (large arrow), which are seen curving toward the bone, still cover the articular surface of the tendon. The greater tuberosity surface is irregular (small black arrows). Such lesions cannot be seen on arthroscopy.

the diagnosis of rotator cuff tears. Similar results were found by Farin et  al.64 In our prospective study of rotator cuff disease, all patients with fluid in the bursa had a rotator cuff tear.13

Joint Effusion Joint fluid can be found in the joint recesses, including the infraspinatus, subcoracoid, and axillary recesses. In a patient who sits in the upright position, most fluid will accumulate in the dependent portion of the biceps tendon sheath. Approximately half of these effusions are associated with rotator cuff tears.6 The other half result from a variety of articular causes of shoulder disease. When a large fluid collection is found in the infraspinatus recess without fluid in the subdeltoid bursa, inflammatory or infectious causes of joint disease should always be excluded.41

896   PART III  ■  Small Parts, Carotid Artery, and Peripheral Vessel Sonography

Bursa

Bursa B

s s

A

B

FIGURE 22-23.  Subdeltoid bursal effusion. A, Transverse view over the anterior shoulder demonstrates fluid in two different synovial compartments. Synovial effusion (s) surrounds the long biceps tendon (B). This fluid does not extend beyond the biceps groove. The larger collection of fluid noted deep to the deltoid fills the subdeltoid bursa and extends both medial and lateral to the confines of the groove. B, Longitudinal scan of the long biceps tendon shows joint effusion (s) extending deep to the tendon. The subdeltoid bursa extends as a large sac over the anterior aspect of the shoulder. Fluid in joint and bursa signifies rotator cuff tear in most patients.

Concave Subdeltoid Fat Contour In the normal shoulder, the bright linear echoes from the subdeltoid bursal fat are convex. Concavity of the subdeltoid contour may be noted in medium and large tears, reflecting the absence of cuff tendon. It may be possible to approximate the deltoid and the humeral surface even in smaller tears using transducer compression at the site of the tear.

Bone Surface Irregularity Only recently has bone irregularity been cited in imaging literature as an important and common associated finding in rotator cuff tears.8,10,58,65 The majority of partial-thickness and full-thickness tears of the distal 1 cm of the rotator cuff are associated with small bone spurs and pits in the bone surface of the greater tuberosity. Use of higher-frequency transducers for rotator cuff imaging may have made these findings more evident. The tuberosity abnormality matches the tendon abnormality in location, size, and shape. The cause of the abnormality is unknown. Trauma from an impaction of the tuberosity on the acromion during shoulder elevation has been considered.

Tear Size and Muscle Atrophy Several studies have attempted to quantify rotator cuff tears. Ultrasound is capable of measuring tears as accurately as MRI if the tears are relatively small. However, both MRI and ultrasound tend to underestimate the size of tears compared with measurement at surgery. If the diameter of the tear exceeds 3 cm, ultrasound assessment is more limited.66,67 New emphasis has been placed on quantifying the muscle loss that accompanies chronic

tears of the rotator cuff. Sofka et al.68 demonstrated that fatty atrophy shows as increased echogenicity in muscles with torn tendons. The teres minor tendon was mentioned as the only muscle that occasionally atrophies without being torn. Newer methods of estimating atrophy have used criteria that assess the surface of the tendon and visibility of the pennate structure and central aponeurosis.69,70 The most reproducible measurement thus far has been the “occupation ratio,” which can be assessed on sagittal images at the level of the suprascapular notch medial to the acromial process.70

Pathology of Rotator Cuff Interval The subscapularis, superior glenohumeral, and coracohumeral ligaments can be distinguished from the conjoint tendons of the rotator cuff. High-frequency transducer technology currently available allows the diagnosis of hyperemia and fibrosis, seen as a mass within the interval of patients with adhesive capsulitis.71 In diabetic patients in particular, the differential diagnosis occasionally must be made between rotator cuff tears and adhesive capsulitis. Tears of the rotator cuff interval and abnormalities of the subscapularis have also been diagnosed with more confidence recently. Detection of these tears is important because these defects require an open approach different from the arthroscopic surgical approach of the more common tears of the supraspinatus and infraspinatus tendons.72,73

POSTOPERATIVE APPEARANCES The literature suggests that sonography can play an important role in the postoperative follow-up after rotator cuff repair.74,75 Because surgery may distort sono-

Chapter 22  ■  The Rotator Cuff   897

graphic landmarks, sonography in the postoperative patient is more difficult than in the preoperative patient. It is therefore important to understand the surgical procedures used in acromioplasty and cuff repair. In acromioplasty the anterior inferior aspect of the acromion is surgically removed. Sonographically, this appears as disruption of the normal, rounded, smooth acromial contour. After surgery, the acromion appears pointed (Fig. 22-24). Because the inferior aspect of the acromion is removed, a greater extent of the supraspinatus tendon may be visualized. Repair of a cuff tear creates unique sonographic landmarks. The cuff tendons are reimplanted into a trough

A

D

GT

D

SUP

B FIGURE 22-24.  Rotator cuff repair. A, Drawing demonstrates the surgical technique for cuff reimplantation, with creation of trough (arrow) in the humeral head, reimplantation of the residual tendon within that trough, and characteristic method of suture placement. B, Longitudinal supraspinatus tendon (SUP) image shows characteristic appearances of reimplantation trough (arrows). Acromioplasty defect (open arrow) is also visualized. D, Deltoid muscle; GT, greater tuberosity; curved arrow, reimplantation suture. (From Mack LA, Nyberg DA, Matsen FA 3rd, et al. Sonography of the postoperative shoulder. AJR Am J Roentgenol 1988;150:1089-1093.)

made perpendicular to the axis of the supraspinatus tendon. The reimplantation trough is placed in the humerus at a site that provides optimal tendon tension. The trough appears sonographically as a defect in the humeral contour, which is best viewed with the transducer longitudinal to the supraspinatus tendon (Fig. 22-25, A), with the shoulder in extension. Suture material may be seen deep in the trough as specular echoes. Scanning the arm in extension and internal rotation may be necessary to visualize this site of tendon reimplantation, especially when it is medially placed (Fig. 22-25, B). Failure to scan in this position may lead to a falsepositive diagnosis. Such a maneuver, however, should be used with care, especially in the immediate postoperative period, to avoid reinjury of the friable, newly reimplanted tendons. Recent improvements in arthroscopic repair have culminated in a more anatomic tendon reconstruction. The two-row repair shows as two reflective surfaces at the placement of the anchors, best seen on longi­ tudinal images. One anchor is typically placed close to the articular margin (medial anchor) and one more laterally in the tuberosity.76 The number of anchors is proportionate to the size of the original defect in the cuff. Sonographic appearances of the cuff tendons never return to normal in the postoperative patient. Tendons, especially the supraspinatus, are often echogenic and thinned when compared with the contralateral shoulder. Joint effusions are common and best visualized along the biceps tendon. Because resection of the subdeltoid bursa removes an important landmark, dynamic scanning is especially important in distinguishing a thin, hyperechoic cuff from adjacent deltoid muscle. In patients with shoulder arthroplasty, ultrasound can be used to demonstrate tears that develop postoperatively.77 These tears can explain postoperative shoulder pain in some patients. The subscapularis tears in particular are important lesions to detect because they can lead to further deterioration of the shoulder function through anterior instability.78

Recurrent Tear Sonographically, recurrent tears most often appear as absence of the cuff. Fluid filling a defect in a rotator cuff repair and loose sutures or screws are other indications of recurrent tear (Fig. 22-26). Unless baseline scans are available in the postoperative period, it may be difficult to differentiate small, recurrent tears from the appearances created when only a small amount of cuff tendon remains to be reattached. Thinning of the tendon is useless as a criterion, and bone irregularity is the rule in the postoperative patient. Recurrent tears are common, occurring in up to 40% of patients with repair of a small defect and 80% of patients with large tears preoperatively.

898   PART III  ■  Small Parts, Carotid Artery, and Peripheral Vessel Sonography

D

D AS GT

SUP GT

A

B

FIGURE 22-25.  Postoperative rotator cuff: importance of examination during extension. A, Longitudinal supraspinatus tendon view in neutral position of a patient after repair of full-thickness rotator cuff tear demonstrates the reimplantation trough, but fails to reveal evidence of the supraspinatus tendon, thus suggesting recurrent injury. B, Scan with the arm in extension and internal rotation demonstrates that the repair is intact. The residual supraspinatus tendon (SUP) is thinned. Note absence of characteristic echoes of the subdeltoid bursa (arrowheads). AS, Acromial shadow; D, deltoid muscle; GT, greater tuberosity; arrows, reimplantation trough. (From Mack LA, Nyberg DA, Matsen FA. Sonographic evaluation of rotator cuff. Radiol Clin North Am 1988;25:161-177.)

Deltoid

Humerus

FIGURE 22-26.  Recurrent tear: postoperative ultrasound examination. Longitudinal scan along the deltoid muscle in the region of the reimplantation trough (arrow). A loose suture (small arrows) is noted within the subdeltoid bursal effusion. The supraspinatus has left this subdeltoid space. The proximal humerus has an abnormal, round appearance. The anatomic neck has disappeared through the process of bone remodeling.

PITFALLS IN INTERPRETATION Inadequate transducer positioning is the most common error in scanning the rotator cuff. False-positive and false-negative results may be produced in this manner. For example, scanning the supraspinatus tendon transversely with the transducer placed laterally may artifactually mimic a rotator cuff tear. An oblique transverse scan of the supraspinatus tendon can be falsely reported as

thinning of the cuff. The examiner must therefore view the cuff in two orthogonal planes. Visualization of neatly depicted bony contour will help in avoiding these pitfalls. A cause of tendon heterogeneity is the geometric relationship of the tendon to the transducer. As demonstrated by Crass et al.79 and Fornage,80 failure to orient the transducer parallel to the fibers of the tendon may result in artifactual areas of decreased echogenicity (Fig. 22-27). When only a small area of the tendon is parallel to the transducer, a focal area of increased echogenicity may be produced, mimicking a small, partial-thickness or full-thickness tear. This artifact is especially pronounced with sector transducers.

ROTATOR CUFF CALCIFICATIONS Calcifications can affect any of the four tendons of the rotator cuff. Subscapular tendon calcifications can be particularly difficult to diagnose without the aid of ultrasound. The calcium can burst out from the tendon into the subacromial-subdeltoid bursa and cause an acute and very painful inflammatory synovitis.81 Standard texts on calcific tendinitis have distinguished a chronic phase of formation and an acute phase of resorption.82 Ultrasound appears incapable of staging calcium according to these phases. However, increases in color Doppler signal have been noted in more painful calcifications.83 Ultrasound has also shown great potential in demonstrating the physical form of the crystal deposition.84 Aggregates of calcium can be solid, pastelike, or liquid (Fig. 22-28). The liquid deposits appear hyperechoic without shadow; the calcium paste casts a vague shadow; and hard deposits show with distinct acoustic shadow. This unique capability of ultrasound aids in the treatment when ultrasound is used to localize and aspirate calcium.85,86

Chapter 22  ■  The Rotator Cuff   899

A

B

FIGURE 22-27.  Artifactual areas of decreased tendon echogenicity. A and B, Two views of the same supraspinatus tendon demonstrate considerable changes in echogenicity that may be artifactually created by transducer position and orientation.

D

S S

Bi B

A

C

B

FIGURE 22-28.  Rotator cuff calcifications. A, Longitudinal image of the supraspinatus (S) shows striated areas of hypoechogenicity (arrow) in the supraspinatus insertion; D, deltoid. B, Transverse view confirms the abnormal echogenicity (arrows) within substance of the supraspinatus (S); Bi, intracapsular biceps. There is no distinct shadowing behind these deposits. C, Shoulder radiograph confirms that this hyperechogenicity represents calcification (arrow).

900   PART III  ■  Small Parts, Carotid Artery, and Peripheral Vessel Sonography

Before the procedure, the clinician will decide on the size and number of needles necessary for treatment. Expectations for complete recovery are lower if the pain is caused by rock-type calcifications. A small amount of corticosteroid is then often added after multiple passes have been made through the calcium, using 16- to 18-gauge needles. In a retrospective study of 44 patients at a minimum of 8 months’ follow-up, 75% of patients noted statistically significant improvement after the lavage of calcium under ultrasound guidance.87 The authors noted the need for randomized clinical trials on this subject, in part because the natural history of calcific tendinitis can result in spontaneous healing. When all other treatment of calcifications fails, calcium deposits can be marked under ultrasound guidance to add precision to the surgical evacuation of the lesions.88

References 1. Matsen FA, Arntz CT. Subacromial impingement. In Matsen FA, editor. The shoulder. Philadelphia: Saunders; 1990. 2. Neviaser RJ, Neviaser TJ. Observations on impingement. Clin Orthop Relat Res 1990:60-63. 3. Resnick D. Shoulder arthrography. Radiol Clin North Am 1981;19: 243-253. 4. Mack LA, Matsen 3rd FA, et al. Ultrasound evaluation of the rotator cuff. Radiology 1985;157:205-209. 5. Mack LA, Gannon MK, Kilcoyne RF, Matsen 3rd RA. Sonographic evaluation of the rotator cuff: accuracy in patients without prior surgery. Clin Orthop Relat Res 1988:21-27. 6. Middleton WD, Reinus WR, Totty WG, et al. Ultrasonographic evaluation of the rotator cuff and biceps tendon. J Bone Joint Surg Am 1986;68:440-450. 7. Crass JR, Craig EV, Feinberg SB. Ultrasonography of rotator cuff tears: a review of 500 diagnostic studies. J Clin Ultrasound 1988; 16:313-337. 8. van Holsbeeck MT, Kolowich PA, Eyler WR, et al. Ultrasound depiction of partial-thickness tear of the rotator cuff. Radiology 1995;197:443-446. 9. Dinnes J, Loveman E, McIntyre L, Waugh N. The effectiveness of diagnostic tests for the assessment of shoulder pain due to soft tissue disorders: a systematic review. Health Technol Assess 2003;7:iii, 1-166. Clinical Considerations 10. Codman EA. The shoulder. 2nd ed. Boston: Thomas Todd; 1934. 11. DePalma AF. Surgery of the shoulder. 2nd ed. Philadelphia: Lippincott; 1973. 12. Refior HJ, Krodel A, Melzer C. Examinations of the pathology of the rotator cuff. Arch Orthop Trauma Surg 1987;106:301-308. 13. Milgrom C, Schaffler M, Gilbert S, van Holsbeeck M. Rotator cuff changes in asymptomatic adults: the effect of age, hand dominance and gender. J Bone Joint Surg Br 1995;77:296-298. 14. Sher JS, Uribe JW, Posada A, et al. Abnormal findings on magnetic resonance images of asymptomatic shoulders. J Bone Joint Surg Am 1995;77:10-15. 15. Raven PB. Asymptomatic tears of the rotator cuff are commonplace. Sports Med Diagn 1995;17:11-12. 16. Miniaci A, Dowdy PA, Willits KR, Vellet AD. Magnetic resonance imaging evaluation of the rotator cuff tendons in the asymptomatic shoulder. Am J Sports Med 1995;23:142-145. 17. Petterson G. Rupture of the tendon aponeurosis of the shoulder joint in antero-inferior dislocation. Acta Chir Scand Suppl 1942;77:1187. 18. Harvie P, Ostlere SJ, Teh J, et al. Genetic influences in the aetiology of tears of the rotator cuff: sibling risk of a full-thickness tear. J Bone Joint Surg Br 2004;86:696-700. 19. Neer Jr CS. Anterior acromioplasty for the chronic impingement syndrome in the shoulder: a preliminary report. J Bone Joint Surg Am 1972;54:41-50.

Technical Considerations 20. Churchill RS, Fehringer EV, Dubinsky TJ, Matsen FA 3rd. Rotator cuff ultrasonography: diagnostic capabilities. J Am Acad Orthop Surg 2004;12:6-11. 21. Strobel K, Zanetti M, Nagy L, Hodler J. Suspected rotator cuff lesions: tissue harmonic imaging versus conventional ultrasound of the shoulder. Radiology 2004;230:243-249. 22. Al-Shawi A, Badge R, Bunker T. The detection of full-thickness rotator cuff tears using ultrasound. J Bone Joint Surg Br 2008;90: 889-892. 23. Kwon D, Bouffard JA, van Holsbeeck M, et al. Battling fire and ice: remote guidance ultrasound to diagnose injury on the International Space Station and the ice rink. Am J Surg 2007;193:417420. 24. Fincke EM, Padalka G, Lee D, et al. Evaluation of shoulder integrity in space: first report of musculoskeletal ultrasound on the International Space Station. Radiology 2005;234:319-322. Anatomy and Sonographic Technique 25. Lee HS, Joo KB, Park CK, et al. Sonography of the shoulder after arthrography (arthrosonography): preliminary results. J Clin Ultrasound 2002;30:23-32. 26. Taljanovic MS, Carlson KL, Kuhn JE, et al. Sonography of the glenoid labrum: a cadaveric study with arthroscopic correlation. AJR Am J Roentgenol 2000;174:1717-1722. 27. Schydlowsky P, Strandberg C, Galatius S, Gam A. Ultrasonographic examination of the glenoid labrum of healthy volunteers. Eur J Ultrasound 1998;8:85-89. 28. Schydlowsky P, Strandberg C, Tranum-Jensen J, et al. Post-mortem ultrasonographic assessment of the anterior glenoid labrum. Eur J Ultrasound 1998;8:129-133. 29. Schydlowsky P, Strandberg C, Galbo H, et al. The value of ultrasonography in the diagnosis of labral lesions in patients with anterior shoulder dislocation. Eur J Ultrasound 1998;8:107-113. 30. Cole BJ, Rodeo SA, O’Brien SJ, et al. The anatomy and histology of the rotator interval capsule of the shoulder. Clin Orthop Relat Res 2001:129-137. 31. Le Corroller T, Cohen M, Aswad R, Champsaur P. [The rotator interval: hidden lesions?]. J Radiol 2007;88:1669-1677. 32. Morag Y, Jacobson JA, Lucas D, et al. Ultrasound appearance of the rotator cable with histologic correlation: preliminary results. Radiology 2006;241:485-491. 33. Rakofsky M. Fractional arthrography of the shoulder. Stuttgart: Gustav Fischer; 1987. 34. Ptasznik R, Hennessy O. Abnormalities of the biceps tendon of the shoulder: sonographic findings. AJR Am J Roentgenol 1995;164: 409-414. 35. Farin PU, Jaroma H, Harju A, Soimakallio S. Medial displacement of the biceps brachii tendon: evaluation with dynamic sonography during maximal external shoulder rotation. Radiology 1995;195: 845-848. 36. van Holsbeeck M, Strouse PJ. Sonography of the shoulder: evaluation of the subacromial-subdeltoid bursa. AJR Am J Roentgenol 1993; 160:561-564. 37. Crass JR, Craig EV, Feinberg SB. The hyperextended internal rotation view in rotator cuff ultrasonography. J Clin Ultrasound 1987; 15:416-420. 38. Ferri M, Finlay K, Popowich T, et al. Sonography of full-thickness supraspinatus tears: comparison of patient positioning technique with surgical correlation. AJR Am J Roentgenol 2005;184:180184. 39. Mack LA, Nyberg DA, Matsen FA 3rd. Sonographic evaluation of the rotator cuff. Radiol Clin North Am 1988;26:161-177. 40. Brestas PS, Tsouroulas M, Nikolakopoulou Z, et al. Ultrasound findings of teres minor denervation in suspected quadrilateral space syndrome. J Clin Ultrasound 2006;34:343-347. 41. van Holsbeeck M, Introcaso J, Hoogmartens M. Sonographic detection and evaluation of shoulder joint effusion. Radiology 1990;15: 416-420. The Normal Cuff 42. Newman JS, Adler RS, Bude RO, Rubin JM. Detection of soft tissue hyperemia: value of power Doppler sonography. AJR Am J Roentgenol 1994;163:385-389. 43. Petersson CJ. Ruptures of the supraspinatus tendon: cadaver dissection. Acta Orthop Scand 1984;55:52-56.

Chapter 22  ■  The Rotator Cuff   901 44. Farin P, Jaroma H. Sonographic detection of tears of the anterior portion of the rotator cuff (subscapularis tendon tears). J Ultrasound Med 1996;15:221-225. 45. Keyes EL. Observations on rupture of the supraspinatus tendon: based upon a study of seventy-three cadavers. Ann Surg 1933;97: 849-856. 46. Wilson CL, Duff G. Pathologic study of degeneration and rupture of the supraspinatus tendon. Arch Surg 1943;47:121-135. 47. Fukuda H, Mikasa M, Yamanaka K. Incomplete thickness rotator cuff tears diagnosed by subacromial bursography. Clin Orthop Relat Res 1987:51-58. 48. Yamaguchi K, Tetro AM, Blam O, et al. Natural history of asymptomatic rotator cuff tears: a longitudinal analysis of asymptomatic tears detected sonographically. J Shoulder Elbow Surg 2001;10: 199-203. 49. Brasseur JL, Lucidarme O, Tardieu M, et al. Ultrasonographic rotator cuff changes in veteran tennis players: the effect of hand dominance and comparison with clinical findings. Eur Radiol 2004;14:857864. Preoperative Appearances 50. Teefey SA, Middleton WD, Payne WT, Yamaguchi K. Detection and measurement of rotator cuff tears with sonography: analysis of diagnostic errors. AJR Am J Roentgenol 2005;184:1768-1773. 51. Middleton WD, Payne WT, Teefey SA, et al. Sonography and MRI of the shoulder: comparison of patient satisfaction. AJR Am J Roentgenol 2004;183:1449-1452. 52. Le Corroller T, Cohen M, Aswad R, et al. Sonography of the painful shoulder: role of the operator’s experience. Skeletal Radiol 2008;37: 979-986. 53. Middleton WD, Teefey SA, Yamaguchi K. Sonography of the rotator cuff: analysis of interobserver variability. AJR Am J Roentgenol 2004;183:1465-1468. 54. O’Connor PJ, Rankine J, Gibbon WW, et al. Interobserver variation in sonography of the painful shoulder. J Clin Ultrasound 2005;33: 53-56. 55. Middleton WD. Status of rotator cuff sonography. Radiology 1989; 173:307-309. 56. Bloom RA. The active abduction view: a new maneuver in the diagnosis of rotator cuff tears. Skeletal Radiol 1991;20:255-258. 57. van Holsbeeck M, Introcaso JH, Kolowich PA. Sonography of tendons: patterns of disease. Instr Course Lect 1994;43:475-481. 58. Jiang Y, Zhao J, van Holsbeeck MT, et al. Trabecular microstructure and surface changes in the greater tuberosity in rotator cuff tears. Skeletal Radiol 2002;31:522-528. 59. Teefey SA, Middleton WD, Bauer GS, et al. Sonographic differences in the appearance of acute and chronic full-thickness rotator cuff tears. J Ultrasound Med 2000;19:377-378; quiz 383. 60. Sorensen AK, Bak K, Krarup AL, et al. Acute rotator cuff tear: do we miss the early diagnosis? A prospective study showing a high incidence of rotator cuff tears after shoulder trauma. J Shoulder Elbow Surg 2007;16:174-180. 61. Jacobson JA, Lancaster S, Prasad A, et al. Full-thickness and partialthickness supraspinatus tendon tears: value of ultrasound signs in diagnosis. Radiology 2004;230:234-242. 62. Guerini H, Feydy A, Campagna R, et al. [Harmonic sonography of rotator cuff tendons: are cleavage tears visible at last?]. J Radiol 2008; 89:333-338. 63. Hollister MS, Mack LA, Patten RM, et al. Association of sonographically detected subacromial/subdeltoid bursal effusion and intraarticular fluid with rotator cuff tear. AJR Am J Roentgenol 1995;165:605608. 64. Farin PU, Jaroma H, Harju A, Soimakallio S. Shoulder impingement syndrome: sonographic evaluation. Radiology 1990;176:845-849. 65. Wohlwend JR, van Holsbeeck M, Craig J, et al. The association between irregular greater tuberosities and rotator cuff tears: a sonographic study. AJR Am J Roentgenol 1998;171:229-233. 66. Bryant L, Shnier R, Bryant C, Murrell GA. A comparison of clinical estimation, ultrasonography, magnetic resonance imaging,

and arthroscopy in determining the size of rotator cuff tears. J Shoulder Elbow Surg 2002;11:219-224. 67. Kluger R, Mayrhofer R, Kroner A, et al. Sonographic versus magnetic resonance arthrographic evaluation of full-thickness rotator cuff tears in millimeters. J Shoulder Elbow Surg 2003;12:110-116. 68. Sofka CM, Haddad ZK, Adler RS. Detection of muscle atrophy on routine sonography of the shoulder. J Ultrasound Med 2004;23:10311034. 69. Strobel K, Hodler J, Meyer DC, et al. Fatty atrophy of supraspinatus and infraspinatus muscles: accuracy of ultrasound. Radiology 2005; 237:584-589. 70. Khoury V, Cardinal E, Brassard P. Atrophy and fatty infiltration of the supraspinatus muscle: sonography versus MRI. AJR Am J Roentgenol 2008;190:1105-1111. 71. Lee JC, Sykes C, Saifuddin A, Connell D. Adhesive capsulitis: sonographic changes in the rotator cuff interval with arthroscopic correlation. Skeletal Radiol 2005;34:522-527. 72. Flury MP, John M, Goldhahn J, et al. Rupture of the subscapularis tendon (isolated or in combination with supraspinatus tear): when is a repair indicated? J Shoulder Elbow Surg 2006;15:659-664. 73. Lyons RP, Green A. Subscapularis tendon tears. J Am Acad Orthop Surg 2005;13:353-363. Postoperative Appearances 74. Mack LA, Nyberg DA, Matsen 3rd FR, et al. Sonography of the postoperative shoulder. AJR Am J Roentgenol 1988;150:10891093. 75. Crass JR, Craig EV, Feinberg SB. Sonography of the postoperative rotator cuff. AJR 1988;148:561-564. 76. Anderson K, Boothby M, Aschenbrener D, van Holsbeeck M. Outcome and structural integrity after arthroscopic rotator cuff repair using 2 rows of fixation: minimum 2-year follow-up. Am J Sports Med 2006;34:1899-1905. 77. Westhoff B, Wild A, Werner A, et al. The value of ultrasound after shoulder arthroplasty. Skeletal Radiol 2002;31:695-701. 78. Sofka CM, Adler RS. Sonographic evaluation of shoulder arthroplasty (original report). AJR Am J Roentgenol 2003;180:1117-1120. Pitfalls in Interpretation 79. Crass JR, van de Vegte GL, Harkavy LA. Tendon echogenicity: ex vivo study. Radiology 1988;167:499-501. 80. Fornage BD. The hypoechoic normal tendon: a pitfall. J Ultrasound Med 1987;6:19-22. Rotator Cuff Calcifications 81. Resnick D, Niwayama G. Diagnosis of bone and joint disorders. 2nd ed. Philadelphia: Saunders; 1988. 82. Gartner J, Simons B. Analysis of calcific deposits in calcifying tendinitis. Clin Orthop Relat Res 1990:111-120. 83. Chiou HJ, Chou YH, Wu JJ, et al. Evaluation of calcific tendonitis of the rotator cuff: role of color Doppler ultrasonography. J Ultrasound Med 2002;21:289-295; quiz 296-297. 84. Farin PU. Consistency of rotator cuff calcifications: observations on plain radiography, sonography, computed tomography, and at needle treatment. Invest Radiol 1996;31:300-304. 85. Farin PU, Jaroma H, Soimakallio S. Rotator cuff calcifications: treatment with ultrasound-guided technique. Radiology 1995;195:841843. 86. Chiou HJ, Chou YH, Wu JJ, et al. The role of high-resolution ultrasonography in management of calcific tendonitis of the rotator cuff. Ultrasound Med Biol 2001;27:735-743. 87. Lin JT, Adler RS, Bracilovic A, et al. Clinical outcomes of ultrasoundguided aspiration and lavage in calcific tendinosis of the shoulder. HSS J 2007;3:99-105. 88. Kayser R, Hampf S, Seeber E, Heyde CE. Value of preoperative ultrasound marking of calcium deposits in patients who require sur­ gical treatment of calcific tendinitis of the shoulder. Arthroscopy 2007;23:43-50.

CHAPTER 23 

The Tendons Bruno D. Fornage, Didier H. Touche, and Beth S. Edeiken-Monroe

Chapter Outline ANATOMY INSTRUMENTATION AND SONOGRAPHIC TECHNIQUE NORMAL SONOGRAPHIC APPEARANCE Shoulder Elbow Hand and Wrist Knee Foot and Ankle

PATHOLOGY Tears

Complete Tears Incomplete Tears

Tendinosis Inflammation

Tendinitis Peritendinitis Tenosynovitis Bursitis Enthesopathy

The tendons of the extremities are particularly well suited for sonographic examination using high-frequency transducers (up to 20 MHz) because of their superficial location. Also, tendons are best evaluated dynamically during their gliding motion, for which the unique realtime capability of sonography is invaluable. Sonography of the musculoskeletal system in general and of the tendons of the extremities in particular continues to grow in popularity, with tendon sonography now extensively performed by rheumatologists, orthopedic surgeons, physiatrists, and sports medicine physicians. This has led to the replication of many early studies performed by radiologists. Sonography has become the first-line imaging modality in many centers specializing in musculoskeletal imaging and sports medicine worldwide, even where magnetic resonance imaging (MRI) is available. Indeed, in expert hands, high-frequency sonography, combined with physical examination and plain radiography, can solve many diagnostic challenges, making MRI unnecessary. The vast majority of tendon disorders are related to trauma and inflammation and are associated with athletic or occupational activities that result in overuse of the tendon, mostly through excessive tension or repetitive microtrauma.

ANATOMY Tendons are made of dense connective tissue and are extremely resistant to traction forces.1 The densely 902

Nonarticular Osteochondroses Impaired Tendon Motion and Entrapment Postoperative Patterns Tumors and Pseudotumors OTHER IMAGING MODALITIES

packed collagen fibers are separated by a small amount of ground substance with a few elongated fibroblasts and are arranged in parallel bundles. The peritenon is a layer of loose connective tissue that wraps around the tendon and sends intratendinous septa between the bundles of collagen fibers. In large tendons, blood and lymphatic vessels course with nerve endings in these septa, whereas small tendons are almost avascular. At the musculotendinous junction, muscle fibers interdigitate with collagen fibrils. The bony insertion of tendons is usually calcified and characterized by cartilaginous tissue. Tendons usually attach to tuberosities, spinae, trochanters, processes, or ridges. Blood supply to tendons is poor, and nutritional exchange occurs mostly through the ground substance. With aging, the ground substance and fibroblasts decrease, whereas the fibers and fat in the tendon increase. In certain areas of mechanical constraint, tendons are associated with additional structures that provide mechanical support, protection, or both. Fibrous sheaths keep certain tendons close to the bones and prevent them from “bowstringing”; examples include the flexor and extensor retinacula in the wrist, the fibrous sheaths (“pulleys”) of the flexor tendons in the fingers, and the peroneal and flexor retinacula in the foot. The sesamoid bones are intended to reinforce tendons’ strength. Synovial sheaths are double-walled tubular structures that surround some tendons; the inner wall of these sheaths is in intimate contact with the tendon, and the two layers are in continuity with each other at both ends and also occasionally through a mesotenon. A

Chapter 23  ■  The Tendons   903

minimal amount of synovial fluid allows the tendon to glide smoothly within its sheath. Large tendons (e.g., patellar, Achilles) lack a synovial sheath and are surrounded instead by a sheath of loose areolar and adipose tissue known as paratenon. Synovial bursae are small, fluid-filled pouches found in particular locations that act as bolsters to facilitate the motion (play) of tendons.

INSTRUMENTATION AND SONOGRAPHIC TECHNIQUE Because of their wider field of view and their better resolution in the near field relative to those of other types of transducers, linear array electronic transducers are the best choice for tendon sonography. Images of exquisite resolution are obtained with the broad-bandwidth (e.g., 5-12 MHz, 7-15 MHz) linear array transducers that are available on current state-of-the-art scanners (Fig. 23-1). Some mechanical transducers of up to 20 MHz are also found on some commercially available scanners. The width of the field of view (FOV) of most highfrequency broadband linear array transducers is limited to about 4 cm. Although most scanners allow splitting of the screen on the monitor to obtain a montage of two contiguous scans, thereby doubling the width of the field of view, the measurements of lesions that straddle the two half screens are inaccurate if the two contiguous views overlap. The image processing technique known as extended–field of view or panoramic imaging allows stretching the FOV width up to 50 to 60 cm. With the accurate measurement of the structures visualized on those extended sonograms, this technique removed a long-standing limitation of real-time sonography and has been particularly effective in musculoskeletal sonography, in which long anatomic segments or lesions are often scanned2,3 (Figs. 23-2 and 23-3).

Real-time spatial compound scanning involves the acquisition of echoes at a given point in an image using multiple different apertures generated by computed beam-steering technology. The images obtained from the multiple lines of sight are compounded in real time. Real-time compound scanning has shown some success in reducing the amount of speckle in the image, making uniform tissue appear more uniform and boundaries more continuous (Fig. 23-4). This may appear beneficial in imaging the fibrillar texture of tendons and reducing the anisotropy artifact. These potential benefits must be carefully weighed against the risks; the unavoidable blurring associated with this technique obscures minute lesions, and the reduction or disappearance of subtle useful artifacts, such as fine trails of shadowing or comettail artifacts, may prevent the detection of tiny reflectors, such as foreign bodies or microcalcifications. Electronic beam steering is available on some highend scanners. This may be useful when the beam from the linear array transducer is not perpendicular to the tendon and needs to be corrected slightly to hit the tendon fibers at a 90-degree angle.4 This technique helps reduce the anisotropy artifact related to the obliquity or concavity of tendons without the blurring associated with real-time spatial compound scanning (Fig. 23-5). In addition, beam steering changes the image format of linear array transducers from rectangular to trapezoidal and thus widens the FOV. Tissue harmonic imaging (THI) is available with high-frequency linear array transducers. Because THI boosts both spatial and contrast resolutions, it can help in confirming the anechoic appearance of minute and deep-seated fluid collections, such as small joint effusions, ganglia, or early acute tenosynovitis, which would otherwise display spurious echoes on fundamental imaging. Color Doppler imaging is now available not only on high-end but also on midrange and even laptop-type

FIGURE 23-1.  Normal patellar tendon. Longitudinal sonogram of the midportion of the patellar tendon using a 5 to 13–MHz linear array transducer shows the fibrillar echotexture of the tendon (arrows).

904   PART III  ■  Small Parts, Carotid Artery, and Peripheral Vessel Sonography

P

T

FIGURE 23-2.  Normal patellar tendon. Longitudinal extended–field of view sonogram shows both insertions (arrows) of the patellar tendon; P, patella; T, tibia.

FIGURE 23-3.  Normal Achilles tendon. Longitudinal extended-field-of-view sonogram shows the entire length of the Achilles tendon (arrowheads) from its origin to its insertion into the calcaneus (C); K, Kager’s fatty triangle; FHL, flexor hallucis longus muscle; S, termination of soleus muscle; T, tibia.

P T

FIGURE 23-4.  Real-time spatial compounding. Real-time spatial compound longitudinal sonogram of the patellar tendon shows the tendon margins well. Note the associated blur. P, Patella; T, tibia.

portable scanners, and it is always good practice to use it when evaluating inflammatory or tumoral conditions. Power Doppler imaging is preferred because of its greater sensitivity in flow detection, especially in light of the low baseline vascularity of tendons. It is important to keep in mind that the color Doppler signals associated with inflammatory conditions of tendons are easily obliterated by even modest pressure exerted with the transducer, or when the tendon is stretched, such as by flexion of the knee for the patellar tendon or dorsiflexion of the foot for the Achilles tendon5 (Figs. 23-6 and 23-7).

Color Doppler imaging has been used to evaluate and quantitate the excursion velocity and gliding characteristics of some tendons in the hand.6-8 Anecdotal reports on the use of ultrasound contrast agents to enhance the visibility of the blood supply to the largest tendons9-11 are of academic interest and probably not clinically significant. Elastography (or elasticity imaging) is the mapping of elasticity of tissues. This can be achieved with MRI or with sonography. Although manufacturers recently commercialized software providing elastograms, thus far

Chapter 23  ■  The Tendons   905

P T

FIGURE 23-5.  Electronic beam steering. Longitudinal sonogram of the patellar tendon using electronic beam steering to achieve a trapezoidal format. This allows the beam to remain perpendicular to the tendon fibers even at the patellar insertion, thus avoiding areas of false hypoechogenicity. P, Patella; T, tibia.

P

P

B

A

FIGURE 23-6.  Effect of examination technique on power Doppler findings: patellar tendinitis. A, Longitudinal sonogram obtained without pressure exerted on the tendon with the transducer shows significant hypervascularity; P, patella. B, Longitudinal sonogram obtained with the usual pressure applied with the transducer shows the nearly complete disappearance of hypervascularity on the power Doppler vascular signals; P, patella.

P

A

P B

FIGURE 23-7.  Effect of examination technique on power Doppler findings: patellar tendinitis. A, Longitudinal sonogram obtained with the knee extended shows substantial hypervascularity; P, patella. B, Longitudinal sonogram obtained with the knee flexed shows the nearly complete disappearance of vascularity seen by the power Doppler signals; P, patella.

these images remain crude, difficult to obtain, and of questionable clinical value. A combination of longitudinal and transverse scans provides a three-dimensional (3-D) approach to tendon examination. Ultrasound scanners capable of 3-D reconstruction of sonograms are commercially available, but no direct benefit of the use of 3-D sonography in the evaluation of superficial tendons has been reported to date (Fig. 23-8).

Once mandatory with the use of 7.5-MHz probes, standoff pads are no longer needed with the use of high– frequency transducers, whose focal zone can be adjusted to the very first millimeters of the scan. However, a thin standoff pad remains useful for evaluating very superficial tendons, such as the extensor tendons of the fingers at the dorsum of the hand, or tendons coursing in regions with an uneven surface, such as the flexor tendons in the fingers12 (Fig. 23-9). A standoff pad is also used to cor-

906   PART III  ■  Small Parts, Carotid Artery, and Peripheral Vessel Sonography

L T

L T

FIGURE 23-8.  Three-dimensional sonographic examination of flexor tendons of fingers in palm. Top left, Reconstructed coronal sonogram shows the flexor tendons (T) of the third and fourth fingers and the companion lumbrical muscles (L). Top right, Volume rendering; bottom left, transverse sonogram; bottom right, longitudinal sonogram.

P1

P2

P3

FIGURE 23-9.  Normal finger. Longitudinal extended-field-of-view sonogram obtained with a thin standoff pad shows the normal superficial and deep flexor tendons (arrows) coursing along the phalanges; P1, first phalanx; P2, second phalanx; P3, third phalanx. Note that the tendons exhibit normal echogenicity only in the segments that are parallel to the linear array transducer; the tendons are falsely hypoechoic in the segments that lie oblique to the beam.

relate the sonographic findings with the palpation findings. This is accomplished by sliding one or two fingers of one hand between the pad and the skin while keeping the transducer in place over the region of interest with the other hand. This palpation under “sonoscopy” allows the clinician to focus during real-time observation with sonography on the region of palpable concern and, conversely, to appreciate the firmness of the sonographic abnormality. When a standoff pad is used, care should be taken to maintain the ultrasound beam strictly perpendicular to the region being examined and avoid artifacts.13 When examining tendons with sonography, the operator should take full advantage of the real-time capability by examining the tendon at rest and during active and passive mobilization through flexion and extension maneuvers.13 A valuable reference for the normal anatomy of the region being examined is obtained by scanning the corresponding area in the contralateral

TENDON SONOGRAPHY: EXAMINATION TECHNIQUE Use linear array transducer. Use highest frequency available. Identify and correct anisotropy-related artifacts (false hypoechogenicity) caused by improper angle of insonation of the tendon. Always combine longitudinal and transverse scans. Check contralateral tendon for reference. Perform dynamic examination during flexion and extension maneuvers. Use power Doppler sonographic imaging.

extremity or region, although the clinician should always consider the possibility of bilateral tendon disorders. Another advantage of real-time sonography is the accurate guidance during interventional procedures. Aspiration of fluid from or injection of drugs or contrast

Chapter 23  ■  The Tendons   907

agent into the fluid-distended synovial sheath of a tendon or an adjacent bursa can be performed safely under ultrasound guidance.14,15

NORMAL SONOGRAPHIC APPEARANCE All normal tendons are echogenic and display a characteristic fibrillar echotexture on longitudinal scans13 (Fig. 23-10). The higher the frequency, the greater the number of visible fibrils. The fine echogenic lines have been shown to correspond to the interfaces between the collagen bundles and the endotenon.16 No specific sonographic appearance seems to correlate with areas of tendon fragility, the so-called vulnerable zones, where ruptures occur most frequently, such as the area of the Achilles tendon located 2.5 to 6 cm from its insertion into the calcaneus. Although easily seen when surrounded by hypoechoic muscles, tendons are less well demarcated when they are surrounded by echogenic fat. A key step in the identification of tendons is their mobilization under real-time sonographic monitoring on longitudinal scans. On transverse sonograms, the reflective bundles of fibers give rise to a finely punctate echogenic pattern (Fig. 23-11). Transverse scans provide the most accurate measurements of tendon thickness.13 However, because of the small size of the structures measured, meticulous care in the measurement technique must be taken; substantial interobserver variability has been reported.17 Like tendons, nerves are echogenic with a fibrillar echotexture. However, the hypoechoic bundles of axons are thicker than the bundles of fibrils, and at high frequencies, fewer interfaces are seen within a nerve than within a tendon of the same caliber. On transverse sonograms, this results in a honeycomb pattern for nerves and slightly decreased overall echogenicity compared with tendons (Fig. 23-12). Sesamoid bones appear as hyperreflective structures associated with acoustic shadowing (Fig. 23-13). At high frequencies, synovial sheaths appear as thin, hypoechoic underlining of the tendon (Fig. 23-14). The largest synovial bursae (e.g., deep infrapatellar, retrocalcaneal) can be seen on sonograms as flat, hypoechoic structures that contain only a sliver of fluid and are only a few millimeters thick18 (Fig. 23-15). Optimal display of the echogenic fibrillar texture of a tendon requires that the ultrasound beam be strictly perpendicular to the tendon’s axis. The slightest obliquity causes scattering of the beam, which results in an artifactual hypoechogenicity19 referred to as the anisotropic property of tendons (Fig. 23-16). Early erroneous descriptions of hypoechoic normal tendons were caused by this anisotropy artifact. When a linear array transducer is used, the artifact occurs wherever the tendon or a tendon segment is not parallel to the trans-

ducer’s footprint. Rocking the transducer by pressing more firmly on one end usually suffices to bring the footprint of the probe back in a direction parallel to the tendon’s axis. When the anisotropy artifact is caused by a tendon’s curved (concave or convex) course, straightening the tendon through muscle contraction usually clears the artifact (Fig. 23-16). If this is not possible, the alternative is to examine the tendon segment by segment, changing the position of the probe so its footprint is parallel to the segment of the tendon being examined. Transverse scans are equally affected by the tendon anisotropy artifact, with falsely hypoechoic sections being displayed whenever the transverse scan plane is not perpendicular to the tendon’s axis (Fig. 23-16, G and H).

Shoulder Sonography of the rotator cuff and the rest of the shoulder is discussed in Chapter 22.

Elbow The anterior and lateral aspects of the elbow are best examined with the elbow extended. The common extensor tendon, which includes tendons from the extensor digitorum, extensor digiti minimi, extensor carpi ulnaris, and extensor carpi radialis brevis muscles, inserts into the lateral aspect of the lateral epicondyle (Fig. 23-17). Similarly, a common tendon of origin for the superficial flexor muscles, which include the pronator teres, flexor carpi radialis, palmaris longus, flexor carpi ulnaris, and flexor digitorum superficialis muscles, inserts into the medial epicondyle. At the anterior aspect of the extended elbow, the tendon of the biceps brachii muscle can be visualized as it inserts into the radial tuberosity. Because of the oblique direction of that tendon, it usually appears slightly hypoechoic (Fig. 23-18). The cubital bursa, which is located between the tendon and the radial tuberosity to facilitate the tendon’s gliding, is normally not seen. With the elbow flexed at a 90-degree angle, the tendon of the triceps brachii muscle is readily identifiable on both longitudinal and transverse scans as it inserts onto the olecranon (Fig. 23-19).

Hand and Wrist In the carpal tunnel the echogenic tendons of the flexor digitorum profundus (FDP) and flexor digitorum superficialis (FDS) muscles are surrounded by the hypoechoic ulnar bursa and are best seen when the wrist is moderately flexed. The median nerve courses outside the ulnar bursa and anterior to the flexor tendons of the second finger20 (Fig. 23-20, A). On transverse scans, the flexor tendons are seen to move during contraction of the fist. The median nerve is also subject to marked

908   PART III  ■  Small Parts, Carotid Artery, and Peripheral Vessel Sonography

A

C

D

H

I

FIGURE 23-10.  Longitudinal sonograms of normal tendons. All tendons exhibit a fibrillar echotexture, with more interfaces being visualized with higher-frequency transducers. A, Tendon of the long biceps (arrows) at the anterior aspect of the shoulder. B, Tendon of the flexor pollicis longus (arrows) in the thenar area. C, Pair of superficial and deep flexor tendons (arrows) of the third finger in the palm. D, Pair of superficial and deep flexor tendons (arrows) of the third finger at the metacarpophalangeal joint obtained with a 20-MHz transducer. Note the higher number of interfaces depicted within the tendons compared with image C, which was obtained at 13 MHz. E, Patellar tendon scanned at 7.5 MHz. F, Patellar tendon scanned at 13 MHz shows more internal interfaces than in image E. G, Longitudinal scan of Achilles tendon using a 5-MHz transducer shows the echogenic tendon (arrows) with few internal interfaces; F, Kager’s fatty triangle; FHL, flexor hallucis longus muscle; S, standoff pad; T, tibia. H, Achilles tendon scanned at 13 MHz. The fibrillar echotexture of the tendon (arrows) is much better depicted than in image G. I, Tendon of the flexor hallucis longus muscle (arrows) in the distal sole of the foot.

changes in shape at various degrees of flexion of the wrist and fingers as it is deformed and displaced by the moving flexor tendons. The median nerve is slightly less echogenic than the tendons and, as with other major peripheral nerve trunks, appears to comprise multiple hypoechoic tubules, with their interfaces responsible for the median nerve’s overall low echogenicity21 (Fig. 23-20, B). In the palm the pairs of FDP and FDS tendons are clearly identified. On longitudinal scans, the play of the

tendons of a given finger is appreciated in real time during flexion and extension of that finger. On transverse scans, the pairs of FDP and FDS tendons appear as rounded echogenic structures adjacent to the corresponding hypoechoic lumbrical muscles (Fig. 23-21). In the fingers the flexor tendons follow the concavity of the phalanges and therefore are affected by the anisotropy artifact on longitudinal scans along most of their course, except for the segments strictly perpendicular to

Chapter 23  ■  The Tendons   909

M

M B

A

FIGURE 23-11.  Transverse sonograms of normal tendons. A, Transverse scan of the palm of the hand shows the normal echogenic, rounded superficial and deep flexor tendons of the second and third fingers (arrows) adjacent to the hypoechoic lumbrical muscles (curved arrows); M, metacarpal bone. B, Transverse sonogram of the thenar region shows the echogenic round cross section of the tendon of the flexor pollicis longus muscle (arrow) surrounded by the hypoechoic muscles; M, metacarpal bone.

A

B

FIGURE 23-12.  Normal median nerve. A, Longitudinal sonogram of the volar aspect of the forearm shows the mostly echogenic nerve (arrows) between the flexor digitorum superficialis and the flexor digitorum profundus muscles. B, Transverse sonogram shows the typical honeycomb pattern (arrows) that differentiates nerves from tendons. C, Transverse sonogram of the carpal tunnel shows the echogenic cross sections of the flexor tendons (arrows) and the cross section of the median nerve (arrowheads), which is less echogenic than the tendons.

C

S

M

FIGURE 23-13.  Tendons of the foot. Longitudinal sonogram of the medial aspect of the sole of the foot shows the tendon of the flexor hallucis longus muscle (arrowheads) and a sesamoid bone (S); M, first metatarsal bone.

910   PART III  ■  Small Parts, Carotid Artery, and Peripheral Vessel Sonography

FIGURE 23-14.  Synovial sheath of flexor tendons of third finger in palm (15-MHz transducer). A, Longitudinal sonogram shows the echogenic superficial (S) and deep (D) flexor tendons (short arrows) with a typical fibrillar texture; long arrows indicate synovial sheath. B, Transverse scan shows the echogenic cross section of the superficial (S) and deep (D) tendons; arrows indicate synovial sheath.

the FDS tendon, which inserts onto the middle phalanx25 (Fig. 23-23).

Knee

FIGURE 23-15.  Normal infrapatellar bursa. Longitudinal scan of the knee shows the deep infrapatellar bursa (arrows) posterior to the distal patellar tendon (P); T, tibia.

the ultrasound beam22,23 (see Figs. 23-9 and 23-16, E and F ). Some of the fibrous sheaths (pulleys) that maintain the flexor tendons in place and prevent them from bowstringing during flexion of the finger can be visualized on sagittal sonograms as a barely visible, hypoechoic focal thickening of the anterior margin of the flexor tendons (Fig. 23-22). In a cadaver study, sonography demonstrated the A2 (proximal phalanx) pulley in 100% of cases, with a mean length of 16 mm, and the A4 (middle phalanx) pulley in 67% of cases, with a mean length of 6 mm.24 Transverse sonograms of the fingers at the level of the first phalanx can demonstrate the passage of the rounded FDP tendon, which inserts onto the base of the distal phalanx, through the splitting of

Sonography is an excellent technique for visualizing the extensor tendons of the knee.26,27 Because both the quadriceps and the patellar tendons may be slightly concave anteriorly when the knee is extended and the quadriceps relaxed, scans should be obtained during contraction of the quadriceps muscle or with the knee flexed, which straightens the tendons and eliminates the anisotropyrelated artifacts (see Fig. 23-16). The quadriceps tendon comprises the tendons of the rectus femoris, vastus lateralis, vastus medialis, and vastus intermedius muscles, which usually are not distinguished sonographically as separate structures. The quadriceps tendon lies underneath the subcutaneous fat and anterior to a fat pad and the collapsed suprapatellar bursa (Fig. 23-24). On transverse scans, the quadriceps tendon’s cross section is oval. The patellar tendon extends from the patella to the tibial tuberosity over a length of 5 to 6 cm (Fig. 23-25, A). On transverse sections, the patellar tendon has a convex anterior and a flat posterior surface (Fig. 23-25, B). At its midportion, the tendon is about 4 to 5 mm thick and 20 to 25 mm wide.26 The subcutaneous prepatellar and infrapatellar bursae are not normally visible, but the deep infrapatellar bursa may appear as a flattened anechoic structure 2 to 3 mm thick (see Fig. 23-15). Sonography has been used in the evaluation of collateral ligaments of the knee and of the iliotibial band.28,29 Normal ligaments are not always easily Text continued on p. 916.

Chapter 23  ■  The Tendons   911

P

F C

P

F D

E

G

H

F

FIGURE 23-16.  False hypoechogenicity caused by anisotropic property of tendons. A to E, Longitudinal sonograms. A, Sonogram of the distal patellar tendon obtained with a 10-MHz curved array sector transducer. The tendon exhibits normal echogenicity (arrows) only in the narrow midportion of the scan, where the beam is perpendicular to the tendon. On either side, obliquity of the beam is responsible for the tendon’s artifactual hypoechogenicity (open arrows). B, Sonogram of the patellar tendon obtained using the trapezoidal format (electronic beam steering) of a linear array transducer. The beam is perpendicular to the tendon fibers along the entire tendon, resulting in the correct display of the tendon’s echogenicity. C, Sonogram of the quadriceps tendon with knee extended and quadriceps relaxed shows the false hypoechogenicity of the patellar insertion (arrow). D, Sonogram obtained with the knee flexed and quadriceps tendon straightened shows normal echogenicity at the patellar insertion (arrow). E, With the finger fully extended, the flexor tendons are curved and exhibit their normal echogenicity (arrowheads). only in the midportion of the scan. F, Moderate flexion of the joint straightens the tendons, which now display their normal echogenicity along their entire course (arrowheads). G and H, Transverse sonograms. G, Patellar tendon with the scan plane not strictly perpendicular to the tendon’s axis, which results in artifactual hypo­ echogenicity (arrowheads). H, Patellar tendon with scan plane strictly perpendicular to the tendon; normal echogenicity is displayed (arrowheads).

912   PART III  ■  Small Parts, Carotid Artery, and Peripheral Vessel Sonography

FIGURE 23-17.  Normal extensor tendon at elbow. Coronal extended-field-of-view sonogram of the lateral aspect of the elbow shows normal common tendon of the extensor muscles of the forearm at the elbow, with the normal, echogenic tendon (arrowheads) inserting into the lateral epicondyle; H, humerus; R, radius.

R FIGURE 23-18.  Anterior aspect of extended elbow. Longitudinal sonogram shows the oblique biceps tendon (arrows) inserting into the radial tuberosity; R, radial head.

O

H FIGURE 23-19.  Posterior aspect of flexed elbow. Longitudinal sonogram of the tendon of the triceps (arrows); H, humerus; O, olecranon.

Chapter 23  ■  The Tendons   913

A

B FIGURE 23-20.  Flexor tendons of fingers in wrist. A, Longitudinal sonogram of the volar aspect of the wrist shows the median nerve (arrowheads) coursing anterior to the flexor tendons of the index finger (arrows). Note the higher echogenicity of the tendons compared with that of the nerve. B, Transverse sonogram of the wrist in moderate flexion shows the echogenic cross sections of the superficial and deep flexor tendons of the fingers in the hypoechoic ulnar bursa. The arrow points to the oval section of the median nerve.

M

M

FIGURE 23-21.  Superficial and deep flexor tendons of fingers. Transverse sonogram of the palm shows the normal, echogenic, rounded pairs of superficial and deep flexor tendons of the second, third, and fourth fingers (arrows) adjacent to the hypoechoic lumbrical muscles (M).

914   PART III  ■  Small Parts, Carotid Artery, and Peripheral Vessel Sonography

A

B

C FIGURE 23-22.  Tendon pulleys in first and second phalanges of third finger. A, Longitudinal sonogram of the first phalanx of the third finger shows the pulley as a very thin (inframillimetric) hypoechoic band of tissue anterior to the flexor tendons (arrowheads). B, Longitudinal sonogram obtained at 20 MHz of the region indicated with a box on image A shows the distal end of the pulley (arrows). C, Transverse sonogram shows the hypoechoic pulley (arrow).

Chapter 23  ■  The Tendons   915

S D A

C

D S

F FIGURE 23-23.  Relationship between superficial and deep flexor tendons. Transverse sonograms at different levels of the first phalanx of the third finger from the base to the proximal interphalangeal joint. A, Transverse sonogram at the base of the first phalanx shows the superficial tendon (S) above the deep tendon (D). B, The superficial tendon becomes thinner and spreads laterally. C, The superficial tendon has split in two halves (arrows), seen on each side of the round deep tendon, which appears hypoechoic on this scan because of anisotropy. D, The two halves of the superficial tendon have reunited behind the deep tendon. E, The superficial tendon now has the shape of a cup containing the deep tendon, which is now superficial. F, Transverse sonogram obtained at the level of the base of the middle phalanx shows the deep tendon (D) lying anterior to the superficial tendon (S).

P

FIGURE 23-24.  Normal quadriceps tendon. Longitudinal scan shows the echogenic tendon (arrows) surrounded by fat; P, patella.

916   PART III  ■  Small Parts, Carotid Artery, and Peripheral Vessel Sonography

P T

A

B FIGURE 23-25.  Normal patellar tendon. A, Longitudinal extended-field-of-view sonogram shows the tendon from its insertion into the patella (P) to its termination into the anterior tibial tuberosity. Note the prepatellar fibers (arrowhead). T, tibia. B, Transverse sonogram shows the convex anterior and flat posterior surfaces (arrows).

delineated from the articular capsule and the surrounding subcutaneous fatty tissues. In chronic injuries of the medial collateral ligament, sonography can demonstrate calcifications within a thickened hypoechoic ligament; this is known as Pellegrini-Stieda disease.30 A few early reports have claimed good results in the evaluation of the cruciate ligaments.31,32 However, sonographic examination of these tendons is limited because it is virtually impossible to scan them other than obliquely, which results in an artifactual hypoechoic appearance.33 It is therefore difficult to evaluate cruciate tendons other than for gross rupture. As a rule, the cruciate ligaments should be assessed with MRI.

Foot and Ankle The Achilles tendon is formed by the fusion of the aponeuroses of the soleus and gastrocnemius muscles, and it inserts onto the posterior surface of the calcaneus. The Achilles tendon is echogenic and exhibits a characteristic fibrillar texture on longitudinal sonograms.34 The termination of the hypoechoic soleus muscle is easily identified anterior to the origin of the Achilles tendon (Fig. 23-26). The fatty Kager’s triangle, which lies ante-

rior to the distal half of the tendon (see Fig. 23-3), is usually echogenic but may show some individual variation in echogenicity. More anteriorly lie the hypoechoic flexor hallucis longus muscle and the echogenic posterior surface of the tibia. The small, flattened, hypoechoic retrocalcaneal bursa is sometimes seen in the angle formed by the tendon and the calcaneus. The tendon fibers at the bony insertion have a short oblique course, which explains their artifactual hypoechogenicity (Fig. 23-26, C); this appearance should not be mistaken for the subcutaneous calcaneal bursa, which is not normally seen. A sonographic study of the Achilles tendon revealed two tendinous portions of different echogenicity representing the portions arising from the soleus and gastrocnemius muscles.35 On transverse sonograms, the cross section of the Achilles tendon is grossly elliptical and tapers medially. The tendon plane is remarkable in that instead of being strictly coronal, it is slanted anteriorly and medially (Fig. 23-27). Because of this configuration, there is a risk of overestimating the thickness of the tendon on strictly sagittal scans, and measurements should therefore be taken from transverse scans. At 2 to 3 cm superior to its insertion, the Achilles tendon is 5 to 7 mm thick and 12

Chapter 23  ■  The Tendons   917

S

A

B

C

FIGURE 23-26.  Normal Achilles tendon. A, Longitudinal sonogram of the origin of the tendon (arrowheads) shows the termination of the muscular fibers of the soleus muscle (S) that connect to the tendon. B, Longitudinal sonogram of the midportion of the tendon (arrowheads) shows its typical fibrillar echotexture. C, Longitudinal sonogram of the termination of the tendon shows the small retrocalcaneal bursa (arrow) with no fluid.

FIGURE 23-27.  Oblique orientation of normal Achilles tendons. Transverse sonograms of both Achilles tendons of the same subject show the oblique orientation (white lines) of the planes of the tendons, Left, left tendon; right, right tendon.

918   PART III  ■  Small Parts, Carotid Artery, and Peripheral Vessel Sonography

to 15 mm wide.34 A weak positive correlation has been found between the tendon’s thickness and the subject’s height.36,37 In the ankle, sonography demonstrates the tendons of the peroneus longus and brevis muscles laterally and those of the tibialis posterior muscle medially. The tendons of the flexor digitorum longus and flexor hallucis longus muscles can also be identified posterior to the medial malleolus, whereas the tendons of the tibialis anterior, extensor hallucis longus, and extensor digitorum longus muscles are seen at the anterior aspect of the ankle joint.38 Dynamic examination during specific flexion and extension maneuvers of the ankle and foot help identify individual tendons. The ankle tendons are enveloped in synovial sheaths. In a study of ankles of asymptomatic volunteers, a small amount of fluid was found in the posterior tibial and common peroneal tendon sheaths in 71% and 12%, respectively.39 In the foot the examination technique and normal sonographic appearance of the flexor and extensor tendons of the toes do not differ significantly from those of the tendons of the fingers.23

PATHOLOGY Tendon disorders result most often from trauma (tears), noninflammatory degenerative conditions (grouped under the term tendinosis), and inflammatory conditions (tendinitis, peritendinitis).

Tears It is currently acknowledged that most tendon ruptures represent the final stage of progressive destruction of the fibrils. Tears usually occur in tendons that have been rendered fragile by such factors as aging, presence of calcifications, general or local corticosteroid therapy, and underlying systemic diseases (e.g., rheumatoid arthritis, seronegative spondyloarthropathies, lupus erythematosus, diabetes mellitus, gout).40-43

Complete Tears Tears resulting from direct trauma to the tendons (e.g., lacerations) are rare. The vast majority of complete tears result from excessive tension applied to the tendon or from normal tension applied in a movement performed in abnormal conditions. Recent complete tendon tears are often diagnosed clinically. If physical examination is delayed, however, the diagnosis may be indeterminate because of inflammatory changes. Sonography can show the full-thickness discontinuity of the tendon. The gap between the torn tendon fragments is filled with hypoechoic hemorrhagic fluid (or clot) or granulomatous tissue, depending on the age of the lesion (Fig. 23-28). The gap varies in length, and when the torn

SONOGRAPHIC SIGNS OF TENDON TEARS Discontinuity of fibers (partial or complete) Focal thinning of the tendon Hematoma of variable size, usually small Bone fragment (in bone avulsion) Nonvisualization of retracted tendon (in complete tear)

fragments are separated by a long distance, the tendon may not be visualized at all. Nonvisualization of the tendon may occur in complete ruptures of the rotator cuff, biceps brachii tendon, and flexor tendons of the fingers. Excluding ruptures of the Achilles tendon, in which a hematoma can develop around the whole tendon, ruptures are usually associated with minimal focal hemorrhage. With avulsion of the tendon from the bone, one or more bone fragments may appear as bright, echogenic foci with acoustic shadowing.44

Incomplete Tears Accurate sonographic diagnosis of an incomplete tear is important because early diagnosis and treatment will prevent a subsequent complete rupture. However, partial tears are difficult to diagnose clinically and to differentiate from focal areas of tendinosis or tendinitis. Sonographically, recent partial ruptures appear as focal hypoechoic defects with discontinuity of the fibrillar pattern either within the tendon or at its attachment18,45 (Fig. 23-29). A focal irregularity at the tendon’s surface may be the only sign of a small partial tear. A partial rupture may also present as only a focal thinning of the tendon, such as in the rotator cuff. Special mention must be made of intrasubstance tears or splits, which often occur in the ankle tendons and appear as longitudinal hypoechoic clefts46 (Fig. 23-29, B). Subtle sonographic findings may become more apparent on dynamic examination of the tendon during active or passive flexion-extension movements of the associated muscle(s). Three-dimensional evaluation of partial ruptures requires a combination of longitudinal and transverse scans. Indirect signs of tendon rupture include effusion in the tendon sheath or thickening of an adjacent bursa. A sensitivity of 94% has been reported for sonography in the diagnosis of partial tears of the Achilles tendon.45 Other studies have reported the superiority of MRI over sonography in the diagnosis of incomplete Achilles tendon tears.47 Although sonography is an acceptable method for diagnosing complete tears of the Achilles tendon, it is limited in differentiating partial ruptures or even microruptures from focal areas of tendinosis.48,49 In the knee, when the patellar tendon is partially detached from the patellar apex, longitudinal scans show the

Chapter 23  ■  The Tendons   919

A

C FIGURE 23-28.  Complete ruptures involving middle third of Achilles tendon. A, Longitudinal sonogram shows the gap between the two ends (arrows) of the torn tendon, which is filled with echogenic tissue and a minimal amount of fluid. B, Longitudinal sonogram shows the retracted, swollen upper fragment (arrows) surrounded by organizing hematoma. C, Longitudinal sonogram shows the discontinuity of the tendon fibers (arrows).

discontinuity of the tendon fibers, whereas transverse scans obtained inferior to the patellar apex demonstrate the round defect in the midline of the tendon (Fig. 23-29, C). This may be indistinguishable from classic lesions of tendinosis seen at the upper insertion of the tendon.

Tendinosis The term tendinosis is used to describe degenerative changes in a tendon without clinical or histopathologic signs of inflammation within the tendon or paratenon. Most often, it is associated with painful focal or diffuse nodular thickening of the tendon. Tendinosis has been mainly described in the patellar tendon (“jumper’s knee”) and the Achilles tendon (achillodynia). A strong relationship exists between tendinosis and the repetitive microtrauma of overuse injuries. The normal age-related degeneration is probably accelerated with increased stress or decreased resistance of the tendon; this is particularly obvious in sports-related injuries. A wide range of histopathologic changes have been described, including degenerative changes (myxoid and

hyaline degeneration, fibrinoid necrosis, microcysts), regeneration (neovascularization and granulation tissue), and microtears. A constant finding, however, is the absence of inflammatory cells. The clinical distinction between tendinosis and tendinitis is not always straightforward. Sonographically, the lesions of tendinosis appear as focal or diffuse areas of greatly decreased echogenicity and tendon enlargement. In the Achilles tendon the lesions involve preferentially the middle third of the tendon, whereas in the patellar tendon the lesions are most often located at the upper insertion of the tendon. In both locations, however, the tendon can be diffusely swollen with focal or diffuse hypoechoic areas.50 Color and power Doppler ultrasound show increased vascularity, usually from the deep surface of the upper patellar tendon and from the deep surface of the distal Achilles tendon51-53 (Fig. 23-30). Sonography of the patellar tendon showed hypoechoic focal lesions consistent with tendinosis in 14% of asymptomatic athletes with no previous history of jumper’s knee.54 However, the significance of these abnormalities in asymptomatic athletes remains unclear. It was shown in basketball players that

920   PART III  ■  Small Parts, Carotid Artery, and Peripheral Vessel Sonography

A

*

C FIGURE 23-29.  Partial tendon tears. A, Partial rupture of patellar tendon at its insertion into patella. Longitudinal sonogram shows the partial detachment from the patella of the deep fibers of the tendon with a small anechoic hematoma (arrows). B, Partial detachment of superior portion of patellar tendon. Transverse scan shows a well-defined, round, hypoechoic midline hematoma (arrow). The arrowheads indicate the tendon’s margins. C, Posterior tibial tendon split. Coronal sonogram of the ankle shows the central split (*) separating the tendon’s fibers (arrows).

hypoechoic areas in the patellar tendon can resolve, remain unchanged, or increase without predicting symptoms of jumper’s knee.55 In contrast, a study of asymptomatic elite soccer players revealed sonographic abnormalities in 18% of the patellar tendons and 11% of the Achilles tendons; players with abnormal patellar tendons had a 17% risk of developing symptomatic jumper’s knee during the 12-month season, whereas those with abnormal Achilles tendons had a 45% risk of developing Achilles tendinosis.56 Early detection of occult tendinosis should prompt adequate treatment to prevent chronic, therapy-resistant symptoms and subsequent tendon ruptures. Color Doppler sonographic examination of Achilles tendinosis can demonstrate the presence of vessels not only outside but also inside the thickened Achilles tendon, mostly in the ventral portion of the tendon. Ultrasound-guided sclerosis of these vessels has been attempted in the treatment of painful chronic Achilles tendinosis.5,57 Reports that the presence of hypervascu-

larity on color Doppler sonographic imaging is associated with pain58,59 have not been confirmed.60

Inflammation Edema associated with inflammation is responsible for the thickening and decreased echogenicity of the tendons, synovial sheath, or paratenon involved. The increased vascularity associated with inflammation can be depicted with power Doppler sonography,61,62 which can also be used to document response to therapy of patients with inflammatory lesions.63

Tendinitis As with tendinosis, tendinitis may be associated with athletic or occupational activities, but on pathologic examination, there is evidence of acute inflammation, often in addition to preexisting degenerative changes of tendinosis. Tendinitis may affect the whole tendon or

Chapter 23  ■  The Tendons   921

FIGURE 23-30.  Tendinosis. A and B, “Jumper’s knee.” Longitudinal color and power Doppler sonograms show the hypoechoic thickening of the upper third of the patellar tendon with substantial associated hypervascularity. C and D, Achillodynia. Longitudinal sonograms of the Achilles tendon show thickening and decreased echogenicity of the tendon with a minimal but unequivocal increase in vascularity at the deep surface of the tendon.

SONOGRAPHIC SIGNS OF TENDINITIS Thickening of the tendon Decreased echogenicity Blurred margins Increased vascularity on color flow Doppler Calcifications in chronic tendinitis

only part of it. For example, in the patellar tendon, focal tendinitis, like tendinosis, often involves the upper insertion of the tendon, whereas focal involvement of the distal insertion typically occurs after surgical transposition of the tibial tuberosity. Sonographically, in acute tendinitis the tendon is thickened, and the margins are often poorly defined. There is also a diffuse decrease in echogenicity.26,34 Because improper scanning may result in a falsely hypoechoic tendon, the examination technique must be flawless. Power Doppler ultrasound is used to document the focal or diffuse increase in vascularity (Fig. 23-31). Comparison with sonograms of the unaffected contralateral tendons is often useful. The presence of flow in a focal area of decreased echogenicity confirms the diagnosis of focal tendinitis and rules out an acute partial tear, because blood flow is not expected to be present

in the blood-filled cavity resulting from the tear. Power Doppler sonography can also be used to monitor a patient’s response to anti-inflammatory therapy. A decrease in size of the tendon and a return to a normal level of echogenicity and very low vascularity indicate healing. In chronic tendinitis the margins of the tendon may be deformed and bumpy. Sonography can detect minute intratendinous calcifications, which appear as bright foci with or without acoustic shadowing, occasionally with a comet-tail artifact. As a rule, the size and shape of these calcifications are better appreciated on low-kilovoltage radiographs, easily obtained with the use of a mammographic unit64 (Fig. 23-32).

Peritendinitis In peritendinitis the inflammation takes place in the paratenon, the layer of connective tissue that wraps around the tendon in the absence of a synovial sheath. This condition is frequently found in the Achilles tendon. Sonographically, peritendinitis is characterized by a hypoechoic thickening of the peritenon, with the tendon remaining grossly unaffected. Because gray-scale sonography is often unable to diagnose mild peritendinitis with sufficient reliability,48 power Doppler imaging is very helpful in documenting the increased vascularity associated with this condition61 (Fig. 23-33).

922   PART III  ■  Small Parts, Carotid Artery, and Peripheral Vessel Sonography

T

A

C

B

FIGURE 23-31.  Tendinitis. A, Tendinitis of tibial insertion of patellar tendon. Longitudinal power Doppler sonogram shows the focal area of decreased echogenicity and hypervascularity; T, tibia. B, Tendinitis of distal Achilles tendon. Longitudinal sonogram shows the swelling and decreased echogenicity of the tendon; C, calcaneus. C and D, Achilles tendinitis. Longitudinal power and spectral Doppler sonograms show the hypervascularity of the diffusely swollen and hypoechoic tendons.

Tenosynovitis Tenosynovitis is defined as the inflammation of a tendon sheath. Any tendon surrounded by a synovial sheath— especially tendons in the hand, wrist, and ankle—can be affected. Trauma, including repetitive microtrauma, and

pyogenic infection are most often responsible for acute tenosynovitis. Cases of tenosynovitis caused by a foreign body retained within a tendon sheath in the hand have been reported.65 Sonographically, the diagnosis of acute tenosynovitis is made when fluid, even a minimal quantity, is identified in the sheath66-68 (Fig. 23-34). Internal

Chapter 23  ■  The Tendons   923

FIGURE 23-32.  Chronic calcified patellar tendinitis. A, Longitudinal scan of the lower attachment of the tendon shows a markedly thickened, hypoechoic tendon (long arrows) with blurred contours and tiny hyperechoic calcifications (short arrows), one with a comet-tail artifact (arrowhead). B, Lateral low-kilovoltage radiograph obtained with a mammographic unit shows the swollen patellar tendon and the small calcifications (arrow); T, tibia; P, patella.

FIGURE 23-33.  Achilles peritendinitis. Longitudinal power Doppler sonogram shows the hypoechoic thickening of the paratenon (arrows) anterior to the tendon (arrowheads) and the associated increased vascularity.

924   PART III  ■  Small Parts, Carotid Artery, and Peripheral Vessel Sonography

FIGURE 23-34.  Tenosynovitis. A, Mild tenosynovitis of posterior tibial tendon at ankle. Coronal sonogram shows a minimal amount of fluid in the tendon sheath. B, Acute tenosynovitis of flexor digitorum tendon in hand. C, Transverse sonogram of the wrist demonstrates fluid surrounding the flexor tendons. D, Tenosynovitis of peronei tendons. Coronal power Doppler sonogram shows fluid in the synovial sheath and hypervascularity around the tendons.

echoes representing debris can be seen in suppurative tenosynovitis, a serious condition that, if left untreated, can lead to the rapid destruction of the tendon.69 Chronic tenosynovitis is characterized by a hypoechoic thickening of the synovial sheath, most often with little or no fluid (Fig. 23-35). The thickening of the sheath may impair the movement of tendons in narrow passages. In de Quervain’s tenosynovitis the tendons of the abductor pollicis longus and extensor pollicis brevis muscles are constricted by the thickened sheath in the pulley over the radial styloid process. Sonography can demonstrate the hypoechoic thickening of the tendon sheath23,70 (Fig. 23-35, B), and power Doppler sonography may demonstrate increased vascularity in the tissues involved. Sonography can be used to guide the injection of contrast medium into the sheath for tenography, a study that silhouettes the sheath wall but cannot demonstrate its thickness. Sonography has also been used to guide injection of steroids into the synovial sheath of the posterior tibial tendon in patients with chronic inflammatory arthropathy.71

Rheumatoid arthritis has a predilection for synovial tissues, including tendon sheaths in the distal extremities. Sonography has proved effective in the diagnosis of rheumatoid tenosynovitis in the hand.72,73 The tendon sheath involved by the pannus is extremely hypoechoic, and occasionally, fluid is also present in the sheath, which enhances the visibility of the pannus (Fig. 23-36). Power Doppler ultrasound shows significant hypervascularity of the pannus. Sonographic findings of tendon involvement include thickening and nonhomogeneity of the tendon, with margins that appear jagged.74 At a later stage, sonography can demonstrate a marked thinning of the tendon or a partial or complete rupture.75

Bursitis Bursitis most often involves the subdeltoid, olecranal, radiohumeral, patellar, and calcaneal bursae. Trauma and, more importantly, repetitive microtrauma play a major role in bursitis, although in many cases, no initiating factor can be found. Prepatellar bursitis, also known

Chapter 23  ■  The Tendons   925

A

P2 B

P1

as “housemaid’s knee,” is a common finding in subjects who spend extended periods kneeling, such as carpet layers.76 Transient accumulation of fluid in the subacromial bursa has been demonstrated on sonograms of the shoulder for as long as 16 to 20 hours after handball training.77 In the early acute stage of bursitis, when the bursa is filled with fluid, sonograms demonstrate a sonolucent, fluid-filled collection with poorly defined margins. In the chronic stage, a complex sonographic appearance with internal echogenic debris results from the presence of granulomatous tissue, precipitated fibrin, and occasionally calcification. Power Doppler imaging often shows increased vascularity in the thickened wall of the bursa and around it78,79 (Fig. 23-37). Because the

FIGURE 23-35.  Chronic tenosynovitis. A, Chronic tenosynovitis of flexor digitorum tendons after surgical treatment of carpal tunnel syndrome. Longitudinal sonogram of the volar aspect of the wrist shows the thickened hypoechoic bursa (arrows) and the absence of any substantial amount of fluid. B, Chronic posttraumatic tenosynovitis of flexor digitorum tendons of index finger. Longitudinal sonogram shows the hypoechoic thickened synovial sheath (arrows), which contains no fluid. Note the grossly intact flexor tendons, with the superficial tendon inserting into the base of the second phalanx. P1, first phalanx; P2, second phalanx. C, De Quervain’s tenosynovitis. Transverse sonogram of the wrist shows the thickened, hypoechoic synovial sheath (arrows) surrounding the tendons of the abductor pollicis longus and extensor pollicis brevis muscles.

bursa and the adjacent tendon may be involved in the same pathologic process, careful examination of the adjacent tendon is recommended; in 82% of patients with distal third Achilles tendon tendinosis, retrocalcaneal bursitis was also present.79

Enthesopathy Inflammatory enthesopathy, or enthesitis, is defined as an inflammation of the insertion of tendons into the bones. This is usually seen in seronegative spondyloarthropathies, but it can also be occupational, metabolic, drug induced, infective, or degenerative. Tendons usually involved include the patellar and Achilles, as well as the

926   PART III  ■  Small Parts, Carotid Artery, and Peripheral Vessel Sonography

plantar fascia. Sonographically, the tendon insertion appears swollen and hypoechoic, with calcifications developing in chronic lesions, ranging from fine calcifications to bony spurs.80-82 Often, there is coexisting bursitis.

Nonarticular Osteochondroses

FIGURE 23-36.  Rheumatoid tenosynovitis of extensor tendon of finger at dorsum of hand. Transverse scan shows the hypoechoic pannus (arrows) surrounding the tendon (open arrow); M, metacarpal bone.

Osgood-Schlatter and Sinding-Larsen-Johansson diseases are both nonarticular osteochondroses of the knee that occur in ossification centers subjected to traction stress. Both conditions occur in adolescents, typically in boys involved in athletic activities. Although the diagnosis is strongly suggested by the clinical history, radiographic studies are often performed to confirm the diagnosis. High-resolution sonography has been used in the evaluation of these two conditions.83-86 OsgoodSchlatter disease is osteochondrosis of the tibial tuber-

O

A

B

P C FIGURE 23-37.  Bursitis. A, Transverse power Doppler sonogram of the posterior aspect of the elbow shows the thick-walled, fluidcontaining olecranal bursa and the bursa’s hypervascularity. O, olecranon. B, Longitudinal sonogram of the distal arm with the elbow flexed shows the enlarged, hypervascular subtendinous bursa of the triceps brachii muscle. C, Prepatellar bursitis. Longitudinal sonogram shows the fluid-filled subcutaneous prepatellar bursa; P, patella; D, Infrapatellar bursitis. Longitudinal power Doppler sonogram shows the hypervascularity around the distended subcutaneous infrapatellar bursa.

Chapter 23  ■  The Tendons   927

osity. In one study of 70 cases, sonography revealed swelling of the anechoic cartilage in 100%, fragmentation of the echogenic ossification center of the anterior tibial tuberosity in 75%, diffuse thickening of the patellar tendon in 22%, and deep infrapatellar bursitis in 17% of cases85 (Fig. 23-38). Sinding-Larsen-Johansson disease is osteochondrosis of the accessory ossification center at the lower pole of the patella. In this rare disease, sonography can demonstrate the fragmented echogenic ossification center and the swollen hypoechoic cartilage and surrounding soft tissues, including the origin of the patellar tendon.86

Impaired Tendon Motion and Entrapment Sonography has the advantage of showing in real-time the normal and abnormal motion of tendons. The gliding of the extensor pollicis longus tendon has been studied in the wrist.87 In patients with a “snapping” iliopsoas tendon, sonography showed that the snapping was provoked by the sudden flipping of the iliopsoas tendon around the iliac muscle, allowing abrupt contact of the tendon against the pubic bone and producing an audible snap.88 Sonography can confirm in real time the subluxation of the long head of the biceps tendon89 and the peroneal tendons.90 Ultrasound has also shown the entrapment of the flexor tendons of a finger complicating a fracture.91

Postoperative Patterns After surgical repair, tendons remain enlarged, hypoechoic, and heterogeneous with blurred, irregular margins18,92,93 (Fig. 23-39). The internal linear echoes that constitute the tendon’s echotexture are thinner and shorter than in normal tendons. Sonography cannot reliably differentiate recurrent tears and tendinitis from postoperative changes. On postoperative transverse scans, the tendon usually has a rounded cross section. The postoperative pattern may last for several months or even years. Occasionally, sonography can detect bright, echogenic foci caused by residual synthetic suture material or calcification. Postoperative Doppler sonographic studies may demonstrate residual hypervascularization in tendons (Fig. 23-39, C). A long-term follow-up study of ruptured Achilles tendons, most repaired surgically, showed that their average thickness was 12 mm (range, 7-20 mm), compared with 5 mm for the controls, and that 14% of the healed tendons contained calcifications.94 A study comparing the sonographic appearance after surgical repair of Achilles tendon rupture with that after nonsurgical treatment found no difference except for more limited gliding function of the tendon after surgery. In addition, there was a weak correlation between the sonographic findings and the clinical outcome.92

Tumors and Pseudotumors Benign tumors of tendons or their sheaths include giant cell tumors and osteochondromas. The giant cell tumor of tendon sheaths is considered a circumscribed form of pigmented villonodular synovitis. It preferentially involves the flexor surface of the fingers and is usually found in young and middle-aged women. Local recurrences are possible after incomplete excision. Sonographically, giant cell tumors appear as hypoechoic masses, sometimes with lobulated contours.23,95 Power Doppler imaging reveals substantial internal vascularity in 71% of lesions.96 Malignant tumors are rare. Synovial sarcomas may arise from a tendon sheath, appearing as an irregular or lobulated hypoechoic mass, which may contain calcifications. In 95% of patients with familial hypercholesterolemia, sonography demonstrates multiple hypoechoic xanthomas in the Achilles tendon and can detect early focal xanthomas in tendons that are not yet enlarged.97 In 30 adults with familial hypercholesterolemia the mean thickness of the Achilles tendon was 11.1 mm, compared with 4.5 mm in normal subjects and 4.9 mm in a group with nonfamilial hypercholesterolemia.98 The use of a cutoff value of 5.8 mm for the thickness of the Achilles tendon has been reported to yield a sensitivity of 75% and a specificity of 85% for sonography in the diagnosis of familial hypercholesterolemia.99 In familial hypercholesterolemia mutation carriers, sonography increased the clinical diagnosis of xanthomas from 43% to 68%.100 Sonography has also been shown to detect hypoechoic infiltration of the Achilles tendon in 38% of children with familial hypercholesterolemia.101 Sonography can be used to monitor the effect of therapy on the Achilles tendon’s thickness and echotexture. Intratendinous rheumatoid nodules appear on sonograms as hypoechoic nodules.72 In contrast, various appearances have been reported for gouty tophi within or adjacent to tendons. An early report mentioned highly echogenic foci with acoustic shadowing, thus claiming easy differentiation from intratendinous rheumatoid nodules.102 However, another study showed the tophi to be hypoechoic with a peripheral increase in vascularity on color Doppler imaging.103 The sonographic appearances of gouty tophi likely parallel the degree of their calcification and associated inflammation. In dialysis-related amyloidosis, joint synovial membranes and capsules as well as tendons (e.g., supraspinatus) may be thickened, with the amount of thickening increasing with the duration of dialysis.104 Ganglion cysts most often occur in the hand but can develop from any joint or tendon sheath. Sonography demonstrates the oval fluid collection adjacent to the joint space or tendon (Fig. 23-40). Occasionally, chronic cysts have internal echoes, causing the cyst to mimic a hypoechoic solid tumor.

928   PART III  ■  Small Parts, Carotid Artery, and Peripheral Vessel Sonography

FIGURE 23-38.  Osgood-Schlatter disease. Longitudinal power Doppler sonogram shows swelling of the cartilage, fragmentation of the echogenic ossification center of the anterior tibial tuberosity, and deep infrapatellar bursitis.

Baker’s cyst is another type of cyst that often occurs adjacent to a joint. Baker’s cysts are caused by an abnormal distention of the gastrocnemiosemimembranous bursa, which frequently communicates with the knee joint through a slit-shaped opening at the posteromedial aspect of the joint capsule. Baker’s cysts are frequently associated with pathologic conditions that increase the intra-articular pressure through overproduction of synovial fluid, capsular sclerosis, or synovial hypertrophy, most often rheumatoid arthritis. Baker’s cysts present clinically as popliteal masses that can be asymptomatic or symptomatic. Ruptured cysts or large cysts dissecting into the calf produce a swollen, painful limb that mimics thrombophlebitis. A Baker’s cyst typically appears sonographically as a fluid-filled collection.105-107 Occasionally, longitudinal scans demonstrate a second anechoic area anterior to the tendon of the gastrocnemius muscle. Transverse scans confirm that both areas represent sections of the same cyst, which surrounds the tendon of the muscle108 (Fig. 23-41). Internal echoes representing fibrinous strands or debris and synovial thickening can be seen in inflamed or infected cysts. In patients with rheumatoid arthritis, a Baker’s cyst may be completely filled with pannus, thus mimicking a solid mass. Power Doppler sonography demonstrates the hypervascularity of the pannus and differentiates it from debris. Osteochondromatosis can also develop in a Baker’s cyst, giving rise to hyperechoic loose bodies, which cast acoustic shadows when calcified.109 In a recently ruptured cyst, sonography can demonstrate the leak as a subcutaneous fluid collection that extends distally into the lower calf down to the ankle. However, when examination is deferred, the sonographic diagnosis may be more prob-

lematic because the leaking fluid has been resorbed, and only an ill-defined residual hypoechoic area remains (Fig. 23-42).

OTHER IMAGING MODALITIES Although it can silhouette the tendons, particularly when the tendons are surrounded by fat, low-kilovoltage radiography cannot demonstrate their structure. However, plain radiography remains the best modality for unequivocally documenting the presence of fine calcifications in tendons or bursae. Tenography is performed by injecting contrast medium into the tendon’s synovial sheath. This imaging technique provides detailed global views of the inner wall of the sheath but cannot appreciate the thickness of the wall as sonography.110,111 Similarly, bursography consists of direct opacification of a bursa. These two techniques have been replaced by cross-sectional imaging in daily practice. MRI after ultrasound-guided bursography has been used recently to better evaluate the deep and superficial infrapatellar bursae and the radial and ulnar bursae of the wrist.112,113 Computed tomography (CT) has rarely been used in the evaluation of tendons.114,115 MRI, on the other hand, because of its excellent contrast and spatial resolution and multiplanar-imaging capability, has become the modality of choice for soft tissue imaging and the “gold standard” for imaging tendons in the United States.116 However, its cost is about 10 times that of sonography, often for obtaining similar diagnostic information.

Chapter 23  ■  The Tendons   929

A

FIGURE 23-39.  Postoperative patterns. A, Longitudinal scan shows tendon of the palmaris longus muscle after surgical repair of a complete rupture, with focal hypoechoic thickening of the tendon (arrows). The arrowheads indicate normal tendon. B, Longitudinal scan shows patellar tendon 15 months after surgery for tendinitis, with diffusely thickened, heterogeneous, hypoechoic tendon (arrows) with poorly defined margins and minute calcifications (open arrow); P, patella. C, Longitudinal power Doppler sonogram shows the residual thickening and hypervascularity of the upper portion of the patellar tendon. Residual chronic postoperative inflammatory changes in the patellar tendon followed percutaneous fixation of a fracture of the tibial shaft that involved inserting an intramedullary rod through the tendon.

930   PART III  ■  Small Parts, Carotid Artery, and Peripheral Vessel Sonography

M

P1

A

B FIGURE 23-40.  Ganglion cyst. A, Longitudinal sonogram of the first phalanx of the third finger shows a well-defined, 0.4 × 0.2–cm cyst (arrow) anterior to the flexor tendons of the finger (arrowheads). Note the distal acoustic enhancement. M, metacarpal bone; P1, first phalanx. B, Longitudinal view of the wrist demonstrates a small ganglion cyst dorsal to the wrist bones. Note the small neck (arrow) connecting the cyst to the joint.

Chapter 23  ■  The Tendons   931

A

FIGURE 23-41.  Baker’s cyst. A, Longitudinal sonogram shows two fluid collections (arrows) separated by the tendon of the gastrocnemius medialis muscle. B, Transverse sonogram shows that the two collections are parts of the same cyst, which wraps around the tendon of the gastrocnemius medialis muscle. C, Longitudinal sonogram shows a large popliteal cyst.

FIGURE 23-42.  Ruptured Baker’s cyst. Longitudinal extended-field-of-view sonogram of the calf shows a complex mass (arrows) that is connected to a small amount of residual fluid in the popliteal fossa (arrowheads) representing the ruptured cyst.

932   PART III  ■  Small Parts, Carotid Artery, and Peripheral Vessel Sonography

High-frequency sonography is currently the only realtime cross-sectional imaging technique, and it provides unique dynamic information. Sonograms can be quickly obtained along virtually any orientation, and very-highfrequency transducers now provide exquisite spatial and contrast resolution. In experienced hands, in specific anatomic locations, and for specific pathologic conditions (e.g., ankle tendon tears, patellar tendinopathy, epicondylitis), high-resolution sonography has been reported to be almost as accurate as or even more accurate than MRI.117-120 However, because of the small size of the structures being examined and the possibility of significant technique-related artifacts, tendon sonography is operator dependent, requiring skill, adequate training, and sufficient experience to achieve the best results.17,121

References Anatomy 1. McMaster PE. Tendon and muscle ruptures: clinical and experimental studies on the causes and location of subcutaneous ruptures. J Bone Joint Surg 1933;15:705-722. Instrumentation and Sonographic Technique 2. Lin EC, Middleton WD, Teefey SA. Extended field of view sonography in musculoskeletal imaging. J Ultrasound Med 1999;18: 147-152. 3. Fornage BD, Atkinson EN, Nock LF, Jones PH. Ultrasound with extended field of view: phantom-tested accuracy of distance measurements. Radiology 2000;214:579-584. 4. Connolly DJ, Berman L, McNally EG. The use of beam angulation to overcome anisotropy when viewing human tendon with high frequency linear array ultrasound. Br J Radiol 2001;74:183-185. 5. Ohberg L, Lorentzon R, Alfredson H. Neovascularisation in Achilles tendons with painful tendinosis but not in normal tendons: an ultrasonographic investigation. Knee Surg Sports Traumatol Arthrosc 2001;9:233-238. 6. Cigali BS, Buyruk HM, Snijders CJ, et al. Measurement of tendon excursion velocity with colour Doppler imaging: a preliminary study on flexor pollicis longus muscle. Eur J Radiol 1996;23:217-221. 7. Soeters JN, Roebroeck ME, Holland WP, et al. Non-invasive measurement of tendon excursion with a colour Doppler imaging system: a reliability study in healthy subjects. Scand J Plast Reconstr Surg Hand Surg 2004;38:356-360. 8. Oh S, Belohlavek M, Zhao C, et al. Detection of differential gliding characteristics of the flexor digitorum superficialis tendon and subsynovial connective tissue using color Doppler sonographic imaging. J Ultrasound Med 2007;26:149-155. 9. Rudzki JR, Adler RS, Warren RF, et al. Contrast-enhanced ultrasound characterization of the vascularity of the rotator cuff tendon: age- and activity-related changes in the intact asymptomatic rotator cuff. J Shoulder Elbow Surg 2008;17(Suppl):96-100. 10. Zardi EM, Rizzello G, Afeltra A. Color and power Doppler sonography of Achilles tendons before and after contrast medium: what is their role? Ultraschall Med 2007;28:52-56. 11. Koenig MJ, Torp-Pedersen S, Holmich P, et al. Ultrasound Doppler of the Achilles tendon before and after injection of an ultrasound contrast agent: findings in asymptomatic subjects. Ultraschall Med 2007;28:52-56. 12. Fornage BD, Touche DH, Rifkin MD. Small parts real-time sonography: a new “water-path.” J Ultrasound Med 1984;3:355-357. 13. Fornage BD. Ultrasonography of muscles and tendons. Examination technique and atlas of normal anatomy of the extremities. New York: Springer-Verlag; 1988. 14. Sofka CM, Adler RS. Ultrasound-guided interventions in the foot and ankle. Semin Musculoskelet Radiol 2002;6:163-168. 15. Mehdizade A, Adler RS. Sonographically guided flexor hallucis longus tendon sheath injection. J Ultrasound Med 2007;26:233237.

Normal Sonographic Appearance 16. Martinoli C, Derchi LE, Pastorino C, et al. Analysis of echotexture of tendons with ultrasound. Radiology 1993;186:839-843. 17. Brushøj C, Henriksen BM, Albrecht-Beste E, et al. Reproducibility of ultrasound and magnetic resonance imaging measurements of tendon size. Acta Radiol 2006;47:954-959. 18. Fornage BD, Rifkin MD. Ultrasound examination of tendons. Radiol Clin North Am 1988;26:87-107. 19. Fornage BD. The hypoechoic normal tendon: a pitfall. J Ultrasound Med 1987;6:19-22. 20. Jacob D, Cohen M, Bianchi S. Ultrasound imaging of nontraumatic lesions of wrist and hand tendons. Eur Radiol 2007;17: 2237-2247. 21. Silvestri E, Martinoli C, Derchi LE, et al. Echotexture of peripheral nerves: correlation between ultrasound and histologic findings and criteria to differentiate tendons. Radiology 1995;197:291-296. 22. Fornage BD, Rifkin MD. Ultrasound examination of the hand. Radiology 1986;160:853-854. 23. Fornage BD, Rifkin MD. Ultrasonic examination of the hand and foot. Radiol Clin North Am 1988;26:109-129. 24. Hauger O, Chung CB, Lektrakul N, et al. Pulley system in the fingers: normal anatomy and simulated lesions in cadavers at MR imaging, CT, and ultrasound with and without contrast material distention of the tendon sheath. Radiology 2000;217:201-212. 25. McNally EG. Ultrasound of the small joints of the hands and feet: current status. Skeletal Radiol 2008;37:99-113. 26. Fornage BD, Rifkin MD, Touche DH, Segal P. Sonography of the patellar tendon: preliminary observations. AJR Am J Roentgenol 1984;143:179-182. 27. Lee MJ, Chow K. Ultrasound of the knee. Semin Musculoskelet Radiol 2007;11:137-148. 28. De Flaviis L, Nessi R, Leonardi M, Ulivi M. Dynamic ultrasonography of capsulo-ligamentous knee joint traumas. J Clin Ultrasound 1988;16:487-492. 29. Goh LA, Chhem RK, Wang SC, Chee T. Iliotibial band thickness: sonographic measurements in asymptomatic volunteers. J Clin Ultrasound 2003;31:239-244. 30. Brys P, Velghe B, Geusens E, et al. Ultrasonography of the knee. J Belge Radiol 1996;79:155-159. 31. Röhr E. Die sonographische Darstellung des hinteren Kreuzbandes. Röntgenblatter 1985;38:377-379. 32. Scherer MA, Kraus M, Gerngross H, Lehner K. [Importance of ultrasound in postoperative follow-up after reconstruction of the anterior cruciate ligament]. Unfallchirurg 1993;96:47-54. 33. Hsu CC, Tsai WC, Chen CP, et al. Ultrasonographic examination of the normal and injured posterior cruciate ligament. J Clin Ultrasound 2005;33:277-282. 34. Fornage BD. Achilles tendon: ultrasound examination. Radiology 1986;159:759-764. 35. Bertolotto M, Perrone R, Martinoli C, et al. High-resolution ultrasound anatomy of normal Achilles tendon. Br J Radiol 1995;68: 986-991. 36. Koivunen-Niemela T, Parkkola K. Anatomy of the Achilles tendon (tendo calcaneus) with respect to tendon thickness measurements. Surg Radiol Anat 1995;17:263-268. 37. Pang BS, Ying M. Sonographic measurement of Achilles tendons in asymptomatic subjects: variation with age, body height, and dominance of ankle. J Ultrasound Med 2006;25:1291-1296. 38. De Maeseneer M, Marcelis S, Jager T, et al. Sonography of the normal ankle: a target approach using skeletal reference points. AJR Am J Roentgenol 2009;192:487-495. 39. Nazarian LN, Rawool NM, Martin CE, Schweitzer ME. Synovial fluid in the hindfoot and ankle: detection of amount and distribution with ultrasound. Radiology 1995;197:275-278. Pathology 40. Downey DJ, Simkin PA, Mack LA, et al. Tibialis posterior tendon rupture: a cause of rheumatoid flat foot. Arthritis Rheum 1988;31: 441-446. 41. Ismail AM, Balakrishnan R, Rajakumar MK. Rupture of patellar ligament after steroid infiltration: report of a case. J Bone Joint Surg 1969;51B:503-505. 42. Kricun R, Kricun ME, Arangio GA, et al. Patellar tendon rupture with underlying systemic disease. AJR Am J Roentgenol 1980;135: 803-807.

Chapter 23  ■  The Tendons   933 43. Morgan J, McCarty DJ. Tendon ruptures in patients with systemic lupus erythematosus treated with corticosteroids. Arthritis Rheum 1974;17:1033-1036. 44. Kaempffe FA, Lerner RM. Ultrasound diagnosis of triceps tendon rupture: a report of 2 cases. Clin Orthop 1996;332:138-142. 45. Kalebo P, Allenmark C, Peterson L, Sward L. Diagnostic value of ultrasonography in partial ruptures of the Achilles tendon. Am J Sports Med 1992;20:378-381. 46. Waitches GM, Rockett M, Brage M, Sudakoff G. Ultrasonographicsurgical correlation of ankle tendon tears. J Ultrasound Med 1998; 17:249-256. 47. Neuhold A, Stiskal M, Kainberger F, Schwaighofer B. Degenerative Achilles tendon disease: assessment by magnetic resonance and ultrasonography. Eur J Radiol 1992;14:213-220. 48. Paavola M, Paakkala T, Kannus P, Jarvinen M. Ultrasonography in the differential diagnosis of Achilles tendon injuries and related disorders: a comparison between pre-operative ultrasonography and surgical findings. Acta Radiol 1998;39:612-619. 49. Kayser R, Mahlfeld K, Heyde CE. Partial rupture of the proximal Achilles tendon: a differential diagnostic problem in ultrasound imaging. Br J Sports Med 2005;39:838-842. 50. Nicol AM, McCurdie I, Etherington J. Use of ultrasound to identify chronic Achilles tendinosis in an active asymptomatic population. J R Army Med Corps 2006;152:212-216. 51. Hoksrud A, Ohberg L, Alfredson H, Bahr R. Color Doppler ultrasound findings in patellar tendinopathy (jumper’s knee). Am J Sports Med 2008;36:1813-1820. 52. Tan SC, Chan O. Achilles and patellar tendinopathy: current understanding of pathophysiology and management. Disabil Rehabil 2008;30:1608-1615. 53. Leung JL, Griffith JF. Sonography of chronic Achilles tendinopathy: a case-control study. J Clin Ultrasound 2008;36:27-32. 54. Cook JL, Khan KM, Harcourt PR, et al. Patellar tendon ultrasonography in asymptomatic active athletes reveals hypoechoic regions: a study of 320 tendons. Victorian Institute of Sport Tendon Study Group. Clin J Sport Med 1998;8:73-77. 55. Khan KM, Cook JL, Kiss ZS, et al. Patellar tendon ultrasonography and jumper’s knee in female basketball players: a longitudinal study. Clin J Sport Med 1997;7:199-206. 56. Fredberg U, Bolvig L. Significance of ultrasonographically detected asymptomatic tendinosis in the patellar and Achilles tendons of elite soccer players: a longitudinal study. Am J Sports Med 2002;30: 488-491. 57. Ohberg L, Alfredson H. Ultrasound-guided sclerosis of neovessels in painful chronic Achilles tendinosis: pilot study of a new treatment. Br J Sports Med 2002;36:173-175. 58. Cook JL, Kiss ZS, Ptasznik R, Malliaras P. Is vascularity more evident after exercise? Implications for tendon imaging. AJR Am J Roentgenol 2005;185:1138-1140. 59. Reiter M, Ulreich N, Dirisamer A, et al. Colour and power Doppler sonography in symptomatic Achilles tendon disease. Int J Sports Med 2004;25:301-305. 60. Van Snellenberg W, Wiley JP, Brunet G. Achilles tendon pain intensity and level of neovascularization in athletes as determined by color Doppler ultrasound. Scand J Med Sci Sports 2007;17:530-534. 61. Premkumar A, Perry MB, Dwyer AJ, et al. Sonography and MR imaging of posterior tibial tendinopathy. AJR Am J Roentgenol 2002;178:223-232. 62. Richards PJ, Dheer AK, McCall IM. Achilles tendon (TA) size and power Doppler ultrasound (PD) changes compared to MRI: a preliminary observational study. Clin Radiol 2001;56:843-850. 63. Newman JS, Laing TJ, McCarthy CJ, et al. Power Doppler sonography of synovitis: assessment of therapeutic response: preliminary observations. Radiology 1996;198:582-584. 64. Fornage B, Touche D, Deshayes JL, et al. Diagnostic des calcifications du tendon rotulien: comparaison échoradiographique. J Radiol 1984;65:355-359. 65. Howden MD. Foreign bodies within finger tendon sheaths demonstrated by ultrasound: two cases. Clin Radiol 1994;49:419-420. 66. Middleton WD, Reinus WR, Totty WG, et al. Ultrasound of the biceps tendon apparatus. Radiology 1985;157:211-215. 67. Gooding GAW. Tenosynovitis of the wrist: a sonographic demonstration. J Ultrasound Med 1988;7:225-226. 68. García Triana M, Fernández Echevarria MA, Alvaro RL, et al. Pasteurella multocida tenosynovitis of the hand: sonographic findings. J Clin Ultrasound 2003;31:159-162.

69. Jeffrey Jr RB, Laing FC, Schechter WP, et al. Acute suppurative tenosynovitis of the hand: diagnosis with ultrasound. Radiology 1987;162:741-742. 70. Giovagnorio F, Andreoli C, De Cicco ML. Ultrasonographic evaluation of de Quervain disease. J Ultrasound Med 1997;16:685689. 71. Brophy DP, Cunnane G, Fitzgerald O, et al. Technical report: ultrasound guidance for injection of soft tissue lesions around the heel in chronic inflammatory arthritis. Clin Radiol 1995;50:120122. 72. Fornage BD. Soft tissue changes in the hand in rheumatoid arthritis: evaluation with ultrasound. Radiology 1989;173:735-737. 73. Kotob H, Kamel M. Identification and prevalence of rheumatoid nodules in the finger tendons using high-frequency ultrasonography. J Rheumatol 1999;26:1264-1268. 74. Grassi W, Tittarelli E, Blasetti P, et al. Finger tendon involvement in rheumatoid arthritis: evaluation with high-frequency sonography. Arthritis Rheum 1995;38:786-794. 75. Coakley FV, Samanta AK, Finlay DB. Ultrasonography of the tibialis posterior tendon in rheumatoid arthritis. Br J Rheumatol 1994; 33:273-277. 76. Myllymaki T, Tikkakoski T, Typpo T, et al. Carpet-layer’s knee: an ultrasonographic study. Acta Radiol 1993;34:496-499. 77. Kruger-Franke M, Fischer S, Kugler A, et al. [Stress-related clinical and ultrasound changes in shoulder joints of handball players]. Sportverletz Sportschaden 1994;8:166-169. 78. Balint PV, Sturrock RD. Inflamed retrocalcaneal bursa and Achilles tendonitis in psoriatic arthritis demonstrated by ultrasonography. Ann Rheum Dis 2000;59:931-933. 79. Gibbon WW, Cooper JR, Radcliffe GS. Distribution of sonographically detected tendon abnormalities in patients with a clinical diagnosis of chronic Achilles tendinosis. J Clin Ultrasound 2000;28: 61-66. 80. Balint PV, Kane D, Wilson H, et al. Ultrasonography of entheseal insertions in the lower limb in spondyloarthropathy. Ann Rheum Dis 2002;61:905-910. 81. Falsetti P, Acciai C, Lenzi L, et al. Ultrasound of enthesopathy in rheumatic diseases. Mod Rheumatol 2009;19:103-113. 82. Filippou G, Frediani B, Selvi E, et al. Tendon involvement in patients with ochronosis: an ultrasonographic study. Ann Rheum Dis 2008;67:1785. 83. De Flaviis L, Nessi R, Scaglione P, et al. Ultrasonic diagnosis of Osgood-Schlatter and Sinding-Larsen-Johansson diseases of the knee. Skeletal Radiol 1989;18:193-197. 84. Blankstein A, Cohen I, Heim M, et al. Ultrasonography as a diagnostic modality in Osgood-Schlatter disease: a clinical study and review of the literature. Arch Orthop Trauma Surg 2001;121:536539. 85. Bergami G, Barbuti D, Pezzoli F. [Ultrasonographic findings in Osgood-Schlatter disease]. Radiol Med (Torino) 1994;88:368372. 86. Barbuti D, Bergami G, Testa F. [Ultrasonographic aspects of Sinding-Larsen-Johansson disease]. Pediatr Med Chir 1995;17: 61-63. 87. Chen M, Tsubota S, Aoki M, et al. Gliding distance of the extensor pollicis longus tendon with respect to wrist positioning: observation in the hands of healthy volunteers using high-resolution ultrasonography. J Hand Ther 2009;22:44-48. 88. Deslandes M, Guillin R, Cardinal E, et al. The snapping iliopsoas tendon: new mechanisms using dynamic sonography. AJR Am J Roentgenol 190:576, 2008. 89. Armstrong A, Teefey SA, Wu T, et al. The efficacy of ultrasound in the diagnosis of long head of the biceps tendon pathology. J Shoulder Elbow Surg 2006;15:7-11. 90. Neustadter J, Raikin SM, Nazarian LN. Dynamic sonographic evaluation of peroneal tendon subluxation. AJR Am J Roentgenol 2004;183:985-988. 91. Pandey T, Al Kandari SA, Al Shammari SA. Sonographic diagnosis of the entrapment of the flexor digitorum profundus tendon complicating a fracture of the index finger. J Clin Ultrasound 2008; 36:371-373. 92. Moller M, Kalebo P, Tidebrant G, et al. The ultrasonographic appearance of the ruptured Achilles tendon during healing: a longitudinal evaluation of surgical and nonsurgical treatment, with comparisons to MRI appearance. Knee Surg Sports Traumatol Arthrosc 2002;10:49-56.

934   PART III  ■  Small Parts, Carotid Artery, and Peripheral Vessel Sonography 93. Alfredson H, Zeisig E, Fahlström M. No normalisation of the tendon structure and thickness after intratendinous surgery for chronic painful midportion Achilles tendinosis. Br J Sports Med 2009;43:948-949. 94. Bleakney RR, Tallon C, Wong JK, et al. Long-term ultrasonographic features of the Achilles tendon after rupture. Clin J Sport Med 2002;12:273-278. 95. Middleton WD, Patel V, Teefey SA, Boyer MI. Giant cell tumors of the tendon sheath: analysis of sonographic findings. AJR Am J Roentgenol 2004;183:337-339. 96. Wang Y, Tang J, Luo Y. The value of sonography in diagnosing giant cell tumors of the tendon sheath. J Ultrasound Med 2007;26: 1333. 97. Bude RO, Adler RS, Bassett DR, et al. Heterozygous familial hypercholesterolemia: detection of xanthomas in the Achilles tendon with ultrasound. Radiology 1993;188:567-571. 98. Ebeling T, Farin P, Pyorala K. Ultrasonography in the detection of Achilles tendon xanthomata in heterozygous familial hypercholesterolemia. Atherosclerosis 1992;97:217-228. 99. Descamps OS, Leysen X, Van Leuven F, Heller FR. The use of Achilles tendon ultrasonography for the diagnosis of familial hypercholesterolemia. Atherosclerosis 2001;157:514-518. 100. Junyent M, Gilabert R, Zambón D, et al. The use of Achilles tendon sonography to distinguish familial hypercholesterolemia from other genetic dyslipidemias. Arterioscler Thromb Vasc Biol 2005;25: 2203-2208. 101. Koivunen-Niemela T, Viikari J, Niinikoski H, et al. Sonography in the detection of Achilles tendon xanthomata in children with familial hypercholesterolaemia. Acta Paediatr 1994;83:1178-1181. 102. Tiliakos N, Morales AR, Wilson Jr CH. Use of ultrasound in identifying tophaceous versus rheumatoid nodules [letter]. Arthritis Rheum 1982;25:478-479. 103. Gerster JC, Landry M, Dufresne L, Meuwly JY. Imaging of tophaceous gout: computed tomography provides specific images compared with magnetic resonance imaging and ultrasonography. Ann Rheum Dis 2002;61:52-54. 104. Jadoul M, Malghem J, van de Berg B, et al. Ultrasonographic detection of thickened joint capsules and tendons as marker of dialysisrelated amyloidosis: a cross-sectional and longitudinal study. Nephrol Dial Transplant 1993;8:1104-1109. 105. McDonald DG, Leopold GR. Ultrasound B-scanning in the differentiation of Baker’s cyst and thrombophlebitis. Br J Radiol 1972;45:729-732. 106. Gompels BM, Darlington LG. Evaluation of popliteal cysts and painful calves with ultrasonography: comparison with arthrography. Ann Rheum Dis 1982;41:355-359.

107. Strome GM, Bouffard JA, van Holsbeeck M. The knee. In Fornage BD, editor. Musculoskeletal ultrasound. New York: Churchill Livingstone; 1995. p. 201-219. 108. Helbich TH, Breitenseher M, Trattnig S, et al. Sonomorphologic variants of popliteal cysts. J Clin Ultrasound 1998;26:171-176. 109. Moss GD, Dishuk W. Ultrasound diagnosis of osteochondromatosis of the popliteal fossa. J Clin Ultrasound 1984;12:232-233. Other Imaging Modalities 110. Engel J, Luboshitz S, Israeli A, Ganel A. Tenography in de Quervain’s disease. Hand 1981;13:142-146. 111. Gilula LA, Oloff L, Caputi R, et al. Ankle tenography: a key to unexplained symptomatology. Part II. Diagnosis of chronic tendon disabilities. Radiology 1984;151:581-587. 112. Viegas FC, Aguiar RO, Gasparetto E, et al. Deep and superficial infrapatellar bursae: cadaveric investigation of regional anatomy using magnetic resonance after ultrasound-guided bursography. Skeletal Radiol 2007;36:41-46., 113. Aguiar RO, Gasparetto EL, Escuissato DL, et al. Radial and ulnar bursae of the wrist: cadaveric investigation of regional anatomy with ultrasonographic-guided tenography and MR imaging. Skeletal Radiol 2006;35:828-832. 114. Mourad K, King J, Guggiana P. Computed tomography and ultrasound imaging of jumper’s knee: patellar tendinitis. Clin Radiol 1988;39:162-165. 115. Rosenberg ZS, Feldman F, Singson RD, et al. Ankle tendons: evaluation with computed tomography. Radiology 1988;166:221226. 116. Beltran J, Mosure JC. Magnetic resonance imaging of tendons. Crit Rev Diagn Imaging 1990;30:111-182. 117. Rockett MS, Waitches G, Sudakoff G, Brage M. Use of ultrasonography versus magnetic resonance imaging for tendon abnormalities around the ankle. Foot Ankle Int 1998;19:604-612. 118. Nallamshetty L, Nazarian LN, Schweitzer ME, et al. Evaluation of posterior tibial pathology: comparison of sonography and MR imaging. Skeletal Radiol 2005;34:375-380. 119. Warden SJ, Kiss ZS, Malara FA, et al. Comparative accuracy of magnetic resonance imaging and ultrasonography in confirming clinically diagnosed patellar tendinopathy. Am J Sports Med 2007;35:427-436.. 120. Miller TT, Shapiro MA, Schultz E, Kalish PE. Comparison of sonography and MRI for diagnosing epicondylitis. J Clin Ultrasound 2002;30:193-202. 121. O’Connor PJ, Grainger AJ, Morgan SR, et al. Ultrasound assessment of tendons in asymptomatic volunteers: a study of reproducibility. Eur Radiol 2004;14:1968-1973.

CHAPTER 24 

Musculoskeletal Interventions Ronald S. Adler

Chapter Outline TECHNICAL CONSIDERATIONS INJECTION TECHNIQUE INJECTION MATERIALS INJECTION OF JOINTS SUPERFICIAL PERITENDINOUS AND PERIARTICULAR INJECTIONS

Foot and Ankle Hand and Wrist INJECTION OF DEEP TENDONS Biceps Tendon Iliopsoas Tendon BURSAL AND GANGLION CYST INJECTIONS

T he real-time nature of ultrasound makes it ideally suited to provide guidance for a variety of musculo­

skeletal interventional procedures.1-10 Continuous observation of needle position ensures proper placement and allows continuous monitoring of the distribution of injected and aspirated material. The adverse effects of improper needle placement during corticosteroid administration are well documented.11-16 Likewise, decompression of fluid-filled lesions and fragmentation of calcific deposits may be performed. The current generation of high-frequency transducers for small parts sonography allows excellent depiction of soft tissue detail and articular surfaces, particularly in the hand, wrist, foot, and ankle.17 This allows needle placement in non-fluid-distended structures, such as a nondistended joint, tendon sheath, or bursa. The injected agent also produces a contrast effect, which can improve delineation of surrounding structures (e.g., labral morphology) and provide additional information regarding the agent’s distribution.18,19 Ultrasound guidance has broad appeal because it does not involve ionizing radiation; this feature is particularly advantageous in the pediatric population and during pregnancy. Ultrasound-guided injections in the musculoskeletal system include injection of joints, tendon sheaths, bursae, and ganglion cysts. The chapter emphasizes the most common injections performed at my institution, an orthopedic and rheumatology specialty hospital. The most common clinical indication for ultrasound-guided injections generally relates to pain that does not respond to other conservative measures, regardless of the anatomic site. The pain may result from a chronic repetitive injury in the work environment, a sports-related injury,

Calcific Tendinitis INTRATENDINOUS INJECTIONS: PERCUTANEOUS TENOTOMY CONCLUSION

or an underlying inflammatory disorder, such as rheumatoid arthritis.

TECHNICAL CONSIDERATIONS Diagnostic and subsequent interventional examinations are often performed using either linear or curved, phased array transducers, based on depth and local geometry. Needle selection is based on specific anatomic conditions (i.e., depth and size of region of interest). We employ a freehand technique in which the basic principle is to ensure needle visualization as a specular reflector.7 This relies on orienting the needle so that it is perpendicular (or nearly so) to the insonating beam (Fig. 24-1). The needle then becomes a specular reflector, often having a strong ringdown artifact. Although needle guides are available and may be of value, a freehand technique allows greater flexibility in adjusting needle position during a procedure. Further, needle visualization can be enhanced by injecting a small amount of anesthetic and observing the corresponding moving echoes in either gray-scale or color flow sonographic imaging.1 Patient positioning should be assessed first to ensure comfort and optimal visualization of the anatomy. It is important to keep in mind that tendons display inherent anisotropy; they will look hypoechoic if the transducer footprint is not parallel to the tendon.17 Therefore the transducer must be oriented to maximize tendon echogenicity to avoid false interpretation of the tendon as being complex fluid or synovium. An offset may be required at the skin entry point of the needle relative to the transducer to allow for the appropriate needle orientation. Deep 935

936   PART III  ■  Small Parts, Carotid Artery, and Peripheral Vessel Sonography

structures, such as tendons about the hip, are often better imaged using a curved linear or sector transducer, operating at center frequencies of about 3.5 to 7.5 MHz. Superficial, linearly oriented structures, such as in the wrist or ankle, are best approached using a linear array transducer with higher center frequencies (>10 MHz). Transducers with a small footprint (“hockey stick”) are particularly well suited to superficial injections. These factors should be assessed before skin preparation. The immiscible nature of the steroid anesthetic mixture may likewise produce temporary contrast effect (Fig. 24-2). In vitro experiments suggest that this prop-

T N

FIGURE 24-1.  Needle as specular reflector with reverberation artifact. A 25-gauge needle (N) has been positioned into the retrocalcaneal bursa deep to the Achilles tendon (T). Note that the needle is a specular reflector with a characteristic reverberation artifact (arrows).

BASELINE

erty is caused by alterations in acoustic impedance by the scattering material, formed by the suspension of steroid in an aqueous background; this results in an increase in echo intensity of about 20 dB.19 This contrast effect has the advantage of increasing the conspicuity of the delivered agent during real time, enabling the operator to better define the distribution of delivered agent during ultrasound-guided therapy (Video 24-1).

INJECTION TECHNIQUE We employ a sterile technique; the area in question is cleaned with iodine-based solution and draped with a sterile drape (Fig. 24-3). The transducer is immersed into iodine-based solution and surrounded by a sterile drape. A drape is also placed over portions of the ultrasound unit. A sonographer or radiologist positions the transducer; a radiologist positions the needle and performs the procedure. We use 1% lidocaine and bupivacaine (0.25%-0.75%) for local anesthesia. Once the needle is in position, the procedure is undertaken while imaging in real time. Depending on anatomic location, a 1.5-inch or spinal needle with stylet is used to administer the anesthetic-corticosteroid mixture, generally consisting of local long-acting anesthetic and one of the standard injectable corticosteroid derivatives (e.g., triamcinolone). Two approaches to performing injections are long axis and short axis, which relate needle orientation to the structure being injected.9 The long-axis approach refers to needle placement in the plane parallel to the structure

EARLY

LATE

FIGURE 24-2.  Contrast effect. A suspension of anesthetic and triamcinolone has been injected into a cyst phantom. Baseline: Before injection, anechoic “cyst” is shown in a scattering medium, with baseline pixel intensities listed. Early: The early mixing phase is obtained immediately after injection. A contrast effect is evident, in which the cyst becomes almost isoechoic to the background. Late: 20 minutes after injection. In the late phase, apparent gravitational effect results in settling of the suspension toward the dependent portions of the cyst phantom and development of a contrast gradient.

Chapter 24  ■  Musculoskeletal Interventions   937

of interest (Fig. 24-4). For example, longitudinal imaging of the hip to display a hip effusion might be used as the plane to direct the needle for ultrasound-guided aspiration. Alternatively, the short-axis approach refers to needle entry in the plane perpendicular to the long axis of a structure (Fig. 24-5). For example, injection of the retrocalcaneal bursa or metatarsophalangeal joint might use a lateral approach. In our experience, the short-axis approach works well when performing injections or aspirations in small joints and tendon sheaths of the hand and foot. The long-axis approach appears better suited for deep joint injections, such as in the hip or shoulder. It is important to recognize, however, that

FIGURE 24-3.  Sterile technique used for ultrasound-guided interventions. The current setup illustrates a dorsal approach for injection of the first metatarsophalangeal joint. A linear high-frequency transducer with a hockey stick configuration is convenient for injection of small joints, as illustrated here. In this case the needle is in the plane perpendicular to the transducer; the transducer parallels the joint so that the needle is imaged in cross section (short-axis approach).

such approaches serve merely as guidelines, and that no single method necessarily applies to any specific injection.

INJECTION MATERIALS Most injections involve use of a long-acting corticosteroid in combination with a local anesthetic in relatively small volumes. Injectable steroids usually come in either crystalline form, associated with a slower rate of absorption, or a soluble form, characterized by rapid absorption.20-22 Crystalline agents include triamcinolone and methylprednisolone acetate (Depo-Medrol). A common soluble agent is Celestone, which includes a rapidly absorbed betamethasone salt. A reactive inflammatory response or flushing response may occur with crystalline steroids, but typically not with soluble agents.13 The most significant complications associated with injectable steroid use in the musculoskeletal system relate to chondrolysis (when used in weight-bearing joints), depigmentation, fat necrosis, and impaired healing response (when used in soft tissues).11-14 Impaired healing has been associated with tendon, ligament, and plantar fascia rupture. The most frequently used mixtures contain insoluble particles, so a systemic injection could theoretically result in an “embolic phenomenon,” which has been implicated as a mechanism for neurologic complications associated with transforaminal injections. We have not encountered this as a complication when performing injections in the appendicular skeletal system. The most common anesthetics are lidocaine (Xylocaine) and bupivacaine (Marcaine).22,23 Both are characterized as “local injectable anesthetics” but differ in the

PRE-INJECTION

POST-INJECTION

N

fh fn

A

B

FIGURE 24-4.  Long-axis approach: injection of left hip. The long-axis approach is suitable for deep joint injections, such as the hip or shoulder. A, Before injection, 22-gauge spinal needle (N) has been positioned at the femoral head-neck junction in a 50 year-old woman with a labral tear demonstrated on MRI (not shown), to assess relief after therapeutic injection. B, After injection, confirmation of intra-articular deposition of injected material is obtained by the presence of micro-bubbles (arrows) deep to the joint capsule; fh, femoral head; fn, femoral neck.

938   PART III  ■  Small Parts, Carotid Artery, and Peripheral Vessel Sonography PRE-INJECTION

POST-INJECTION C N

N

M

P

A

B

FIGURE 24-5.  Short-axis approach for injection of first metatarsophalangeal (MTP) joint. A, Long-axis view shows 25-gauge needle positioned in MTP joint of 53-year-old woman with plantar plate injury; needle (N) is seen in cross section; M, metatarsal head; P, proximal phalanx. B, While monitoring the injection in real time, the joint capsule distends and fills with echogenic material; C, capsule.

onset of effect and duration. Lidocaine is characterized by early onset (seconds) and short duration (1-2 hours). Bupivacaine becomes effective in 5 to 10 minutes and generally last 4 to 6 hours. In addition to allergic reactions, potential adverse effects include neurotoxicity and cardiotoxicity; these are generally rare when small doses are used under image guidance, taking care to avoid an intravascular injection. Bupivacaine has also been associated with chondrolysis when used for intra-articular applications, but only with constant infusions during arthroscopy and in vitro.24 Chondrolysis is probably not an issue with the small, fixed volumes of bupivacaine typically employed during injections in the musculo­ skeletal system.

INJECTION OF JOINTS A high-frequency linear transducer is used for hand, wrist, elbow, foot and ankle injections. A short-axis approach is often technically easier for small joint injections. The needle should enter the skin parallel to the plane of the joint space. Superficial joints usually appear as separations between the normally continuous specular echoes produced by cortical surfaces. As in other fluidcontaining structures, the presence of an effusion is a helpful feature in visualizing the needle as it enters the joint, because it provides a fluid standoff. The short-axis approach entails scanning across the joint and looking for the transition from one cortical surface to the next, marking the skin (with a surgical marker) and then placing a needle into the joint using ultrasound guidance. When imaging the joint in long axis, the needle will be seen in cross section (Fig. 24-5). Needle placement is confirmed by injecting a small amount of 1% lidocaine, which should display distention of the joint, as well as echoes filling the joint. Small joint injections generally require 0.5 to 1 mL of the therapeutic mixture. In our experience, this approach

C

N

R L

CA

FIGURE 24-6.  Long-axis approach for therapeutic radiocarpal joint injection. A 25-gauge needle (N) has been positioned deep to the dorsal capsule (C) and above the lunate bone (L) of 19-year-old female patient with chronic wrist pain, to assess relief; R, radius; CA, capitate.

works well in the metatarsophalangeal or metacarpophalangeal (MTP/MCP) and interphalangeal (IP) joints, midfoot, ankle, and elbow. Occasionally, a longaxis approach may be efficacious, as in the radiocarpal joint and lateral gutter of the ankle (Fig. 24-6). Ultrasound guidance allows the clinician to negotiate osteophytes and joint bodies. It allows identification of capsular outpouching, thereby affording a more convenient, indirect approach into a joint than slipping a needle into a small joint space. A long-axis approach and a spinal needle are used when performing injections of large joints such as the shoulder or hip (Fig. 24-7). A greater volume is usually injected, typically 5 mL of the steroid-anesthetic mixture. In the case of adhesive capsulitis, significantly larger volumes of local anesthetic (5-10 mL) may be added to provide additional joint distention. We generally approach the glenohumeral joint using a posterior approach, with the patient in a decubitus position and the arm placed in cross-adduction. An intermediatefrequency, linear or curvilinear transducer will suffice in most cases. A linear transducer often results in better

Chapter 24  ■  Musculoskeletal Interventions   939

C D N I

A

CL

AC JT G

H

Left shoulder

FIGURE 24-7.  Long-axis approach for glenohumeral joint injection. A 22-gauge needle (N) has been positioned deep to the posterior capsule (arrows) during a glenohumeral joint injection in 42-year-old woman with adhesive capsulitis. Mild fluid distension of the posterior recess of the joint is evident. H, Humeral head; G, glenoid; I, infraspinatus muscle; D, deltoid muscle.

anatomic detail than curved arrays. The interface of the glenohumeral joint is usually seen with the patient in the decubitus position, as well as the hypoechoic articular cartilage overlying the humeral head. We perform this injection using a long-axis approach, with the needle directed toward the joint along the articular cartilage and deep to the posterior capsule. A test injection with 1% lidocaine should show bright echoes filling the posterior recess or distributed along the articular cartilage (Video 24-1). The hip is approached similarly in long axis, with the transducer placed over the proximal anterior thigh at the level of the joint25 (see Fig. 24-4). The approach is similar to that used in evaluating the joint for an effusion. Ideally, the anterior capsule is imaged at the headneck junction of the femur. In this approach the scan plane is lateral to the neurovascular bundle. The needle may be directed into the joint while maintaining its position in the scan plane of the transducer. A test injection of 1% lidocaine confirms the intra-articular needle position, and the therapeutic injection follows. Fibrous joints, such as the acromioclavicular (AC) joint, can likewise be injected using ultrasound guidance (Fig. 24-8). A short-axis technique is employed similar to that used in the foot. The majority of these injections can be performed using a 1.5-inch needle with a small volume (0.5-1.0 mL) of therapeutic mixture. In addition to the AC joint, this approach is useful in the sternoclavicular joint and pubic symphysis.

SUPERFICIAL PERITENDINOUS AND PERIARTICULAR INJECTIONS Peritendinous injection of anesthetic and long-acting corticosteroid is an effective means to treat tenosynovitis, bursitis, and ganglion cysts in the hand, foot, and ankle.

FIGURE 24-8.  Short-axis approach to injection of acromioclavicular (AC) joint. A 25-gauge needle (long thin arrow) is seen in cross section in a distended hypertrophic AC joint during therapeutic injection in 73-year-old woman with pain centered over AC joint. The joint appears widened, containing echogenic material caused by the contrast effect (short thick arrow) of the therapeutic agent. A, Acromion; CL, clavicle; C, distended capsule.

These structures are superficially located and well delineated on sonography. Ultrasound-guided injections are an effective means to ensure correct localization of therapeutic agents.

Foot and Ankle In my experience, peritendinous injections in the foot and ankle are most often requested for patients with chronic achillodynia or those with medial or lateral ankle pain caused by posterior tibial or peroneal tendinosis or tenosynovitis. Less often, patients are referred to help differentiate pain from posterior impingement and stenosing tenosynovitis of the flexor hallucis longus tendon.26 This distinction can be difficult, sometimes requiring diagnostic and therapeutic injection of the corresponding tendon sheath. Patients with plantar foot pain caused by plantar fasciitis and forefoot pain resulting from painful neuromas are also frequently referred for ultrasound-guided injections.27,28 The large majority of patients with achillodynia have pain referable to the enthesis, with associated retrocalcaneal bursitis and Achilles tendinosis. Enthesis is the site of attachment of a muscle or ligament to bone where the collagen fibers are mineralized and integrated into bone. A retrocalcaneal bursal injection may help alleviate local pain and inflammation (Fig. 24-9). We scan the patient in a prone position with the ankle in mild dorsiflexion, using a linear transducer of 10 MHz or higher frequency. A 1.5-inch needle usually suffices in these patients, with placement using a short-axis approach. The deep retrocalcaneal bursa is usually well seen. A small amount of anesthetic will help confirm position by active distention of the bursa in real time. We similarly approach posterior tibial or peroneal tendons in short axis (Fig. 24-10). Patients with pain in this distribution have been shown to benefit from local

940   PART III  ■  Small Parts, Carotid Artery, and Peripheral Vessel Sonography

T

T N

A

B

C FIGURE 24-9.  Retrocalcaneal bursa injection. A, Short-axis view shows Achilles tendon (T) in 59-year-old man with retrocalcaneal pain and history of Haglund’s deformity. A 25-gauge needle (N) enters perpendicular to the tendon’s long axis and terminates in a small, retrocalcaneal bursal effusion. B, Rotating transducer 90 degrees results in the more typical short-axis view, with the needle (arrow) seen in cross section. C, While observing in real time, the bursa distends (arrows) and fills with echogenic material (contrast effect). The needle is still evident within the distended bursa.

PRE-INJECTION

N

A

POST-INJECTION

T

B

FIGURE 24-10.  Tendon sheath injection using short-axis approach. A 17-year-old female patient with medial ankle pain was referred for ultrasound-guided injection of posterior tibial tendon sheath. A, Preinjection view shows 25-gauge needle (N) within a small, tendon sheath effusion (long arrow) in the inframalleolar portion of the tendon (T). The tendon, which is inhomogeneous, is seen in cross section. B, Postinjection view shows that the tendon sheath is distended, confirming appropriate deposition of the injected material. Note that the tendon margins are better delineated because of a tenosonographic effect of the injected fluid. The vascular pedicle (short arrow) of the tendon is evident.

Chapter 24  ■  Musculoskeletal Interventions   941

tendon sheath injections. The presence of preexisting tendon sheath fluid can facilitate needle visualization. However, careful scanning should be done before the procedure to assess the needle trajectory relative to adjacent neurovascular structures. Use of color or power Doppler sonographic imaging can facilitate visualization of the neurovascular bundle. The posterior tibial nerve is closely related to adjacent vascular structures and is usually well seen before bifurcating into medial and lateral plantar branches. Fluid frequently is seen in relation to the posterior tibial tendon, in the submalleolar region. The peroneal tendons are less predictable. Use of power Doppler sonography in conjunction with realtime guidance can help localize areas of inflammation for guided injection. In stenosing tenosynovitis the tendons may be surrounded only by a thickened retinaculum, proliferative synovium, or scar tissue. In this case, use of a test injection of local anesthesia can be invaluable to confirm the distribution of the therapeutic agent within the tendon sheath in real time. The flexor hallucis longus (FHL) tendon poses a more challenging problem because of its close relation to the neurovascular bundle of the posterior medial ankle. One helpful feature in performing FHL tendon sheath injections is that tendon sheath effusions tend to localize at the posterior recess of the tibiotalar joint. The neurovascular bundle is easily circumvented by placing the needle lateral to the Achilles tendon while scanning medially (Fig. 24-11). This approach allows flexibility in needle placement while maintaining the needle perpendicular to the insonating beam. Ultrasound diagnosis of plantar fasciitis includes thickening of the medial band of the plantar fascia and fat pad edema. One treatment option for severe plantar fasciitis is regional corticosteroid injection, typically performed using anatomic landmarks. However, “blind” injections into the heel have been associated with rupture of the plantar fascia and failure of the longitudinal arch.13 Ultrasound can be used to guide a needle along the

plantar margin of the fascia, thus avoiding direct intrafascial injection.26 The plantar fascia is imaged with the patient prone and the foot mildly dorsiflexed, using a long-axis approach. The transducer is centered over the medial band, which is most often implicated in these patients. A mark is placed over the posterior aspect of the heel and the needle advanced superficial to the plantar fascia, approximately to the margin of the medial tubercle (Fig. 24-12). We perform a perifascial injection using this approach, monitoring the distribution of injected material in real time. Interdigital (Morton’s) neuromas, a common cause of forefoot pain especially in women, have been described at sonography as hypoechoic masses replacing the normal hyperechoic fat in the interdigital web spaces. Occasionally, a dilated hypoechoic tubular structure can be seen associated with the neuroma, reflecting the enlarged feeding interdigital nerve. The second and third web spaces are most often involved. We generally inject Morton’s neuromas using a dorsal approach while imaging the neuroma in long axis28 (Fig. 24-13). This approach is well tolerated by the majority of patients. In certain patients, however, a plantar approach to injecting the nodule is preferred, such as those with severe subluxation at the MTP joint. In either case, the needle is positioned directly within the neuroma and/or adjacent intermetatarsal bursa (if present) and a small volume of therapeutic mixture injected, similar to that used for a small joint injection (0.5 mL).

Hand and Wrist In the hand and wrist, de Quervain’s tendinosis is a frequently encountered tendinopathy involving the abductor pollicis longus and extensor pollicis brevis tendons that responds to local administration of antiinflammatory agents (Fig. 24-14). Injections are also frequently requested for patients with rheumatoid arthritis or psoriatic arthritis. These patients typically

PRE-INJECTION

POST-INJECTION

N TA

A

T

B

FIGURE 24-11.  Flexor hallucis longus (FHL) tendon sheath injection. Short-axis approach with ultrasound guidance in 31-year-old professional dancer with posteromedial ankle pain during plantar flexion. A, Preinjection image depicts the tendon (T) at the level of the posterior sulcus of the talus (TA). The arrows show relationship of the tendon to the neurovascular structures. B, Postinjection image depicts 25-gauge needle (N) situated within the distended tendon sheath (arrows) below the neurovascular structures.

942   PART III  ■  Small Parts, Carotid Artery, and Peripheral Vessel Sonography

experience severe tenosynovitis, which can lead to secondary tendon rupture and deformity. The approach is similar to that used for superficial structures in the foot and ankle. A short-axis approach avoids the surrounding neurovascular structures, and the corresponding tendon sheaths are injected.

N

PF calc

INJECTION OF DEEP TENDONS FIGURE 24-12.  Plantar fascia injection. The proximal medial band of the plantar fascia (PF) is thickened and inhomogeneous (arrows) in a 36-year-old man with hindfoot pain; calc, calcaneus. A 25-gauge needle (N) has been positioned superficial to this plantar fascia and a perifascial injection performed. The injected material (arrows) loculates along the superficial margin of the medial band.

Frequently requested deep tendon injections include those for the bicipital tendon sheath, iliopsoas tendon, gluteal tendon insertion onto the greater trochanter, and hamstring tendon origin.

PRE-INJECTION

POST-INJECTION

N

A

B

FIGURE 24-13.  Morton’s neuroma injection. A, Preinjection image shows 25-gauge needle (N) positioned in a third web space neuroma using a dorsal approach in 45-year-old woman with forefoot pain. Neuroma appears as a heterogeneous hypoechoic nodule (arrows) within the normal echogenic fat. B, After injection and needle removal, the nodule appears expanded and echogenic (arrows). The injected material often decompresses into an adjacent adventitial bursa, which frequently accompanies these nodules.

PRE-INJECTION

POST-INJECTION

N T

ra

A

B

FIGURE 24-14.  Injection of first dorsal compartment of wrist. This 70-year-old woman with de Quervain’s tendinosis had clinical symptoms of wrist pain radiating along the extensor surface of the forearm. A, Preinjection image shows 25-gauge needle (N) positioned in the first dorsal compartment tendon sheath under ultrasound guidance. The tendons (T) are inhomogeneous, with a small effusion evident (arrows) in the dependent part of the tendon sheath. B, After injection and needle removal, the injected material distends the sheath (arrows), producing a tenosonographic effect; the intrinsic tendon abnormalities become more conspicuous; ra, radial artery.

Chapter 24  ■  Musculoskeletal Interventions   943 PRE-INJECTION

POST-INJECTION

N

A

bg

B

bg

FIGURE 24-15.  Biceps tendon sheath injection. Biceps tendinosis is clinically suspected and a biceps tendon sheath injection requested for this 41-year-old man with development of anterior shoulder pain after arthroscopic surgery for labral tear. A, Preinjection image shows 25-gauge needle (N) placed superficial to the long head of the biceps tendon (arrow); bg, bicipital groove. B, After injection and needle removal, there is distension of the tendon sheath by fluid (arrows) containing low-level echoes caused by contrast effect.

Biceps Tendon Anterior shoulder pain with radiation into the arm may be secondary to bicipital tendinitis or tenosynovitis.29 The biceps tendon can be palpated, but if nondistended, the sheath may offer less than 2 mm of clearance to place a needle. This is complicated by the caudal extension of the subacromial subdeltoid bursa, which may overlie the bicipital tendon sheath. A non-image-guided injection could therefore result in delivery into an extratendinous synovial space, or possibly result in an intratendinous injection. We have found that ultrasound guidance enables localization of therapeutic agent to the biceps tendon sheath.10 The patient is placed recumbent with the forearm supinated and the shoulder mildly elevated. The bicipital groove is oriented anteriorly. A linear transducer, typically 7.5 MHz, is used with a lateral approach and 25- or 22-gauge needle (Fig. 24-15). The long head of the biceps tendon is scanned in short axis. When fluid distends the bicipital tendon sheath, the tip is directed into the fluid. Otherwise, the needle is directed along the superficial margin of the tendon, and a test injection of local anesthetic is used to confirm local distention of the sheath, which is then followed by administration of the long-acting corticosteroid. The presence of fluid distention of the sheath with superficially located microbubbles helps to confirm a successful injection.

Iliopsoas Tendon The iliopsoas tendon lies superficial to and along the medial margin of the anterior capsule of the hip. The tendon inserts onto the lesser trochanter. A bursa that frequently communicates with the hip is seen in this location and may be distended because of underling joint pathology or a primary iliopsoas bursitis. Alternatively, iliopsoas tendinosis may occur in the absence of a preexisting bursitis for which a peritendinous injection

is requested.30 A lateral approach to the tendon often requires use of a lower-frequency transducer and curved linear or sector geometry. The neurovascular bundle lies medial and superficial to the tendon, so it is advantageous to approach from the lateral margin of the tendon and perform a small test injection to confirm needle position. A successful injection will show the appearance of fluid or microbubbles distending a bursa that follows the course of the long axis of the tendon (Fig. 24-16).

BURSAL AND GANGLION CYST INJECTIONS Distended bursae around tendinous insertions provide anatomic localization for therapeutic agents. Injection of these areas is often requested for the patient with localized bursitis and abnormality of the adjacent tendon. Examples include the retrocalcaneal, iliopsoas, greater trochanteric, and ischial bursae (Fig. 24-17). Alternatively, the presence of a bursitis, distended synovial cyst, or ganglion cyst may cause mechanical impingement of adjacent tendons. The decompression of these cysts with subsequent administration of a therapeutic agent may alleviate these symptoms31,32 (Fig. 24-18). Ultrasound guidance allows the clinician to avoid intratendinous injections as well as adjacent neurovascular structures. Furthermore, the needle may be redirected as necessary in the presence of a multiloculated cyst (Fig. 24-19).

Calcific Tendinitis The presence of symptomatic intratendinous calcification involves the deposition of calcium hydroxyapatite. This often appears as a nodular echogenic mass within the tendon, which may or may not display posterior acoustic shadowing.33 Although most often affecting the shoulder, this may occur elsewhere in the musculo-

944   PART III  ■  Small Parts, Carotid Artery, and Peripheral Vessel Sonography PRE-INJECTION

fa

POST-INJECTION

fn N T

e

B

A

FIGURE 24-16.  Ultrasound-guided iliopsoas bursa injection for pain relief. This 66-year-old woman with a total hip arthroplasty had developed pain with hip flexion. A, Preinjection image shows 22-gauge spinal needle (N) positioned deep to the tendon (T) at the level of the iliopectineal eminence (e), using a short-axis approach; fa, femoral artery; fn, femoral nerve. B, After injection and needle removal, fluid surrounds the tendon within the distended iliopsoas bursa (arrow).

nv N

fh

e

B

A

FIGURE 24-17.  Ultrasound-guided aspiration and injection of multiloculated iliopsoas bursa. A, Image shows 22-gauge spinal needle (N) positioned into the lateral component of the bursa in 65-year-old woman with groin pain; fh, femoral head; e, iliopectineal eminence. B, After aspiration of the lateral component, the needle has been advanced into the medial component for aspiration and subsequent injection with therapeutic mixture; nv, neurovascular structures. PRE-INJECTION

POST-INJECTION

N C

mhg

A

B

FIGURE 24-18.  Ultrasound-guided aspiration and injection of clinically suspected Baker’s cyst. A, Preinjection image shows 22-gauge needle (N) positioned in the cyst (C) under ultrasound guidance in 59-year-old woman with posterior knee pain and swelling; mhg, medial head of gastrocnemius muscle. B, After cyst aspiration and injection of the therapeutic mixture, the anechoic fluid is replaced by echogenic fluid resulting from contrast effect (arrows).

skeletal system. Ultrasound-guided fragmentation and lavage have been described as an excellent method to reduce the level of calcification and to deposit therapeutic agents.34-37 We currently employ a single-needle technique, with the needle acting as inflow for anesthetic/

sterile saline and as an outflow for the calcium solution (Fig. 24-20). The elasticity of the pseudocapsule encasing the calcification is sufficient to decompress the calcific mass in the majority of cases (Video 24-2). After multiple lavages, the needle is used to inject anesthetic

Chapter 24  ■  Musculoskeletal Interventions   945

c

f

A

N

C

B

FIGURE 24-19.  Ultrasound-guided aspiration and injection of multiloculated ganglion cyst. A, Baseline sonogram shows a multiloculated cyst (c) within the vastus lateralis muscle of the left knee and superficial to the lateral margin of the femur (f ) in 41-year-old woman. B, 20-gauge spinal needle (N) was initially positioned into the proximal component of the cyst; C, Subsequently, the needle was redirected into the distal component. Multiple lavages and aspiration enabled complete decompression of the cyst (not shown).

Right shoulder

Right shoulder

N

D

T

H

A

B

FIGURE 24-20.  Ultrasound-guided aspiration and injection for calcific tendinosis. A, Image shows 20-gauge spinal needle (N) positioned into the calcification (arrow) under ultrasound guidance in 42-year-old man with shoulder pain; H, humeral head; D, deltoid. B, Series of repeat lavage and aspirations of the calcification are performed with the calcification eventually largely replaced by fluid contents within the surrounding pseudocapsule of the calcific mass; T, rotator cuff tendons. Notice that the degree of posterior acoustic shadowing has diminished, and that the center of the calcification (arrow on A) is partially replaced by fluid. After numerous lavages, the calcification is typically fenestrated, and a therapeutic mixture is injected and often decompresses into the subdeltoid bursa (not shown).

and anti-inflammatory mixture. The injected mixture is distributed within the calcification and adjacent subdeltoid bursa in most cases. If the calcification is too small or fragmented, precluding lavage and decompression, the single needle is used to fenestrate the calcium deposit, and a peritendinous therapeutic injection has been shown to be effective.

INTRATENDINOUS INJECTIONS: PERCUTANEOUS TENOTOMY Recent literature suggests that image guidance can be useful for performing percutaneous tenotomy and intratendinous injections with either autologous blood or platelet-rich plasma (PRP).38-42 All these methods are

946   PART III  ■  Small Parts, Carotid Artery, and Peripheral Vessel Sonography

T me

B

N

C

A

FIGURE 24-21.  Ultrasound-guided injection of autologous blood to induce healing response. A, Coronal inversion-recovery MR scan of the affected elbow in 43-year-old man with medial epicondylitis shows increased signal intensity (arrow) of the common flexor tendon mass and adjacent collateral ligament. B, Long-axis ultrasound image of the tendon (T) and adjacent medial epicondyle (me) shows that tendon is predominantly hypoechoic, reflecting underlying tendinosis. C, Image shows 22-gauge needle (N) placed within the common flexor tendon mass, for purposes of mechanical fenestration, and injection of 5 mL of autologous blood, obtained from an antecubital vein. Tendon echogenicity (small arrow) is increased by microbubbles within the injected blood.

T

me

rc

FIGURE 24-22.  Baseline image before autologous blood injection. Common extensor tendon mass (T) in 50-year-old woman with partial tear of the deep portion of the tendon (short arrow, extensor carpi radialis brevis) as it inserts on the medial epicondyle (me); rc, radiocapitellar joint; long arrow, plane of needle entry for percutaneous tenotomy and autologous blood injection. (Same patient is shown in Video 24-3.)

associated with secondary release of local growth factors, such as platelet-derived growth factor (PDGF), which in turn may produce a direct healing response.41 Preliminary data show significant promise in promoting ultrasound-guided tendon repair. “Dry needling” techniques have been employed successfully in patients with lateral epicondylitis refractory to other conservative measures.38 Likewise, autologous blood injections and PRP injec-

tions have been successfully used in both the elbow and the knee39-41 (Figs. 24-21 and 24-22; Video 24-3). The advantage of performing these injections under ultrasound guidance becomes evident when the clinician wants to generalize such techniques to include tendons close to neurovascular structures, such as the hamstring tendon origin.

CONCLUSION Ultrasound offers distinct advantages in providing guidance for delivery of therapeutic injections. Most importantly, ultrasound allows the operator to visualize the needle and make adjustments in real time, to ensure that medication is delivered to the appropriate location. The current generation of ultrasound scanners provides excellent depiction of relevant musculoskeletal anatomy. The needle has a unique sonographic appearance and can be monitored with real-time imaging, as can the steroidanesthetic mixture. Given these advantages, ultrasound guidance should become the method of choice to perform a large variety of guided musculoskeletal interventions.

References 1. Christensen RA, Van Sonnenberg E, Casola G, Wittich GR. Interventional ultrasound in the musculoskeletal system. Radiol Clin North Am 1988;26:145-156.

Chapter 24  ■  Musculoskeletal Interventions   947 2. Cunnane G, Brophy DP, Gibney RG, FitzGerald O. Diagnosis and treatment of heel pain in chronic inflammatory arthritis using ultrasound. Semin Arthritis Rheum 1996;25:383-389. 3. Brophy DP, Cunnane G, Fitzgerald O, Gibney RG. Technical report: ultrasound guidance for injection of soft tissue lesions around the heel in chronic inflammatory arthritis. Clin Radiol 1995;50:120-122. 4. Cardinal E, Chhem RK, Beauregard CG. Ultrasound-guided interventional procedures in the musculoskeletal system. Radiol Clin North Am 1998;36:597-604. 5. Koski JM. Ultrasound-guided injections in rheumatology. J Rheumatol 2000;27:2131-2138. 6. Grassi W, Farina A, Filippucci E, Cervini C. Sonographically guided procedures in rheumatology. Semin Arthritis Rheum 2001;30:347353. 7. Sofka CM, Collins AJ, Adler RS. Use of ultrasonographic guidance in interventional musculoskeletal procedures: a review from a single institution. J Ultrasound Med 2001;20:21-26. 8. Sofka CM, Adler RS. Ultrasound-guided interventions in the foot and ankle. Semin Musculoskelet Radiol 2002;6:163-168. 9. Adler RS, Sofka CM. Percutaneous ultrasound-guided injections in the musculoskeletal system. Ultrasound Q 2003;19:3-12. 10. Adler RS, Allen A. Percutaneous ultrasound-guided injections in the shoulder. Tech Shoulder Elbow Surg 2004;5(2):122-133. 11. Unverferth LJ, Olix ML. The effect of local steroid injections on tendon. J Sports Med 1973;1:31-37. 12. Ford LT, DeBender J. Tendon rupture after local steroid injection. South Med J 1979;72:827-830. 13. Gottlieb NL, Riskin WG. Complications of local corticosteroid injections. JAMA 1980;243:1547-1548. 14. Oxlund H, Manthorpe R. The biochemical properties of tendon and skin as influenced by long-term glucocorticoid treatment and food restriction. Biorheology 1982;19:631-646. 15. Stapczynski JS. Localized depigmentation after steroid injection of a ganglion cyst on the hand. Ann Emerg Med 1991;20:807-809. 16. Shrier I, Matheson GO, Kohl 3rd HW. Achilles tendonitis: are corticosteroid injections useful or harmful? Clin J Sport Med 1996;6:245-250. 17. Bouffard JA, Eyler WR, Introcaso JH, van Holsbeeck M. Sonography of tendons. Ultrasound Q 1993;11:259-286. 18. Koski JM, Saarakkala SJ, Heikkinen JO, Hermunen HS. Use of airsteroid-saline mixture as contrast medium in greyscale ultrasound imaging: experimental study and practical applications in rheumatology. Clin Exp Rheumatol 2005;23:373-378. 19. Luchs JS, Sofka CM, Adler RS. Sonographic contrast effect of combined steroid and anesthetic injections: in vitro analysis. J Ultrasound Med 2007;26:227-231. Injection Materials 20. Curatolo M, Bogduk N. Pharmacologic pain treatment of musculoskeletal disorders: current perspectives and future prospects. Clin J Pain 2001;17:25-32. 21. Caldwell JR. Intra-articular corticosteroids: guide to selection and indications for use. Drugs 1996;52:507-514. 22. Kannus P, Jarvinen M, Niittymaki S. Long- or short-acting anesthetic with corticosteroid in local injections of overuse injuries? A prospective, randomized, double-blind study. Int J Sports Med 1990;11: 397-400. 23. Cox B, Durieux ME, Marcus MA. Toxicity of local anaesthetics. Best Pract Res Clin Anaesthesiol 2003;17:111-136. 24. Gomoll AH, Kang RW, Williams JM, et al. Chondrolysis after continuous intra-articular bupivacaine infusion: an experimental model

investigating chondrotoxicity in the rabbit shoulder. Arthroscopy 2006;22:813-819. Injection of Joints 25. Sofka CM, Saboeiro G, Adler RS. Ultrasound-guided adult hip injections. J Vasc Interv Radiol 2005;16:1121-1123. Superficial Peritendinous and Periarticular Injections 26. Mehdizade A, Adler RS. Sonographically guided flexor hallucis longus tendon sheath injection. J Ultrasound Med 2007;26:233-237. 27. Tsai WC, Wang CL, Tang FT, et al. Treatment of proximal plantar fasciitis with ultrasound-guided steroid injection. Arch Phys Med Rehabil 2000;81:1416-1421. 28. Sofka CM, Adler RS, Ciavarra GA, Pavlov H. Ultrasound-guided interdigital neuroma injections: short-term clinical outcomes after a single percutaneous injection—preliminary results. HSS J 2007;3: 44-49. Injection of Deep Tendons 29. Middleton WD, Reinus WR, Totty WG, et al. Ultrasound of the biceps tendon apparatus. Radiology 1985;157:211-215. 30. Adler RS, Buly R, Ambrose R, Sculco T. Diagnostic and therapeutic use of sonography-guided iliopsoas peritendinous injections. AJR Am J Roentgenol 2005;185:940-943. Bursal and Ganglion Cyst Injections 31. Breidahl WH, Adler RS. Ultrasound-guided injection of ganglia with corticosteroids. Skeletal Radiol 1996;25:635-638. 32. Chiou HJ, Chou YH, Wu JJ, et al. Alternative and effective treatment of shoulder ganglion cyst: ultrasonographically guided aspiration. J Ultrasound Med 1999;18:531-535. 33. Farin PU, Jaroma H. Sonographic findings of rotator cuff calcifications. J Ultrasound Med 1995;14:7-14. 34. Farin PU, Jaroma H, Soimakallio S. Rotator cuff calcifications: treatment with ultrasound-guided technique. Radiology 1995;195: 841-843. 35. Farin PU, Rasanen H, Jaroma H, Harju A. Rotator cuff calcifications: treatment with ultrasound-guided percutaneous needle aspiration and lavage. Skeletal Radiol 1996;25:551-554. 36. Aina R, Cardinal E, Bureau NJ, et al. Calcific shoulder tendinitis: treatment with modified ultrasound-guided fine-needle technique. Radiology 2001;221:455-461. 37. Lin JT, Adler RS, Bracilovic A, et al. Clinical outcomes of ultrasoundguided aspiration and lavage in calcific tendinosis of the shoulder. HSS J 2007;3:99-105. Intratendinous Injections: Percutaneous Tenotomy 38. McShane JM, Nazarian LN, Harwood MI. Sonographically guided percutaneous needle tenotomy for treatment of common extensor tendinosis in the elbow. J Ultrasound Med 2006;25:1281-1289. 39. James SL, Ali K, Pocock C, et al. Ultrasound-guided dry needling and autologous blood injection for patellar tendinosis. Br J Sports Med 2007;41:518-521; discussion 522. 40. Connell DA, Ali KE, Ahmad M, et al. Ultrasound-guided autologous blood injection for tennis elbow. Skeletal Radiol 2006;35:371377. 41. Mishra A, Pavelko T. Treatment of chronic elbow tendinosis with buffered platelet-rich plasma. Am J Sports Med 2006;34:17741778. 42. Gamradt SC, Rodeo SC, Warren RF. Platelet-rich plasma in rotator cuff repair. Tech Orthop 2007;22:26-33.

CHAPTER 25 

The Extracranial Cerebral Vessels Edward I. Bluth and Barbara A. Carroll

Chapter Outline CAROTID ARTERY ANATOMY CAROTID ULTRASOUND EXAMINATION CAROTID ULTRASOUND INTERPRETATION Visual Inspection of Gray-Scale Images Vessel Wall Thickness and IntimaMedia Thickening Plaque Characterization Plaque Ulceration Gray-Scale Evaluation of Stenosis

Doppler Spectral Analysis Standard Examination Spectral Broadening

S

Pitfalls in Interpretation High-Velocity Blood Flow Patterns

Color Doppler Ultrasound

Optimal Settings for Low-Flow Vessel Evaluation Advantages and Pitfalls

Power Doppler Ultrasound Pitfalls and Adjustments

Internal Carotid Artery Occlusion Preoperative Strategies for Patients with Carotid Artery Disease Postoperative Ultrasound

Carotid Artery Stents and Revascularization Grading Carotid Intrastent Restenosis

troke secondary to atherosclerotic disease is the third leading cause of death in the United States. Many stroke victims survive the catastrophic event with some degree of neurologic impairment.1 More than 500,000 new cases of cerebrovascular accident (CVA, stroke) are reported annually.2 Ischemia from severe, flow-limiting stenosis caused by atherosclerotic disease involving the extracranial carotid arteries is implicated in 20% to 30% of strokes.2 An estimated 80% of CVAs are thromboembolic in origin, often with carotid plaque as the embolic source.3 Carotid atherosclerotic plaque with resultant stenosis usually involves the internal carotid artery (ICA) within 2 cm of the carotid bifurcation. This location is readily amenable to examination by sonography as well as surgical intervention. Carotid endarterectomy (CEA) initially proved to be more beneficial than medical therapy in symptomatic patients with carotid stenoses of more than 70%, as reported in the North American Symptomatic Carotid Endarterectomy Trial (NASCET) and the European Carotid Surgery Trial (ECST).4,5 Subsequent NASCET results for moderate stenoses have shown a net benefit for surgical intervention with carotid narrowing between 50% and 69% of vessel diameter. A 15.7% reduction in the 5-year ipsilateral stroke rate was seen in patients treated surgically, versus 22.2% stroke reduction in those treated medically. These results are not as compelling as those for the higher 948

NONATHEROSCLEROTIC CAROTID DISEASE TRANSCRANIAL DOPPLER SONOGRAPHY VERTEBRAL ARTERY Anatomy Sonographic Technique and Normal Examination Subclavian Steal Stenosis and Occlusion INTERNAL JUGULAR VEINS Sonographic Technique Thrombosis

degree of stenosis seen in the earlier NASCET trial. The benefit from surgery was greatest in men, patients with recent stroke, and those with hemispheric symptoms. In addition, the NASCET trials dealing with moderate carotid stenoses required rigorous surgical expertise, such that the risks for disabling stroke or death should not exceed 2% to achieve the statistical surgical benefit.6 The Asymptomatic Carotid Atherosclerosis Study (ACAS) trials published in 1995 reported a reduction in ipsilateral stroke in asymptomatic patients with greater than 60% ICA stenoses who undergo CEA.2 However, these results were less clear-cut than the NASCET trials. Accurate diagnosis of carotid stenosis clearly is critical to identify patients who would benefit from surgical treatment. In addition, ultrasound can assess plaque morphology, such as determining heterogeneous or homogeneous plaque, known to be an independent risk factor for stroke and transient ischemic attack (TIA). Over the past two decades, carotid sonography has largely replaced angiography as the principal screening method for suspected extracranial carotid atherosclerotic disease. Gray-scale examination, color Doppler, power Doppler, and pulsed Doppler imaging techniques are routinely employed in the evaluation of patients with neurologic symptoms and suspected extracranial cerebral disease.7 Ultrasound is an inexpensive, noninvasive, and highly accurate method of diagnosing carotid stenosis.

Chapter 25  ■  The Extracranial Cerebral Vessels   949

INDICATIONS FOR CAROTID ULTRASOUND Evaluation of patients with hemispheric neurologic symptoms, including stroke, transient ischemic attack, and amaurosis fugax. Evaluation of patients with a carotid bruit. Evaluation of pulsatile neck masses. Preoperative evaluation of patients scheduled for major cardiovascular surgical procedures. Evaluation of nonhemispheric or unexplained neurologic symptoms. Follow-up of patients with proven carotid disease. Evaluation of patients after carotid revascularization, including stenting. Intraoperative monitoring of vascular surgery. Evaluation of suspected subclavian steal syndrome. Evaluation of a potential source of retinal emboli. Follow-up of carotid dissection. Follow-up of radiation therapy to the neck in select patients.

Angiography is an expensive, invasive test with potential morbidity, which is why reliance on carotid sonography without preoperative angiography is becoming increasingly common. Magnetic resonance angiography (MRA) and computed tomography (CT) are additional noninvasive screening tools for the identification of carotid bifurcation disease as well as for clarification of ultrasound findings. Angiography is often now reserved for those patients for whom the ultrasound or MRA was equivocal or inadequate. Other carotid ultrasound applications include the evaluation of carotid bruits, monitoring the progression of known atherosclerotic disease,7-9 assessment during or after CEA or stent placement,10 preoperative screening prior to major vascular surgery, and evaluation after the detection of retinal cholesterol emboli.7 Also, nonatherosclerotic carotid diseases can be evaluated, including follow-up of carotid dissection,11-15examination of fibromuscular dysplasia or Takayasu’s arteritis, assessment of malignant carotid artery invasion,16,17 and workup of pulsatile neck masses and carotid body tumors.18,19

CAROTID ARTERY ANATOMY The first major branch of the aortic arch is the innominate or brachiocephalic artery, which divides into the right subclavian artery and right common carotid artery (CCA). The second major branch is the left CCA, which is generally separate from the third major branch, the left subclavian artery (Fig. 25-1). The right and left CCAs ascend into the neck posterolateral to the thyroid gland and lie deep to the jugular vein and sternocleidomastoid muscle. The CCAs have

E

I I

E

V V

C C

S

S In

R

L A

FIGURE 25-1.  Branches of aortic arch and extracranial cerebral arteries. R, Right side; L, left side; A, aortic arch; C, common carotid artery; E, external carotid artery; In, innominate artery; I, internal carotid artery; S, subclavian artery; V, vertebral artery.

different proximal configurations, with the right originating at the bifurcation of the innominate (brachiocephalic) artery into the common carotid and subclavian arteries. The left CCA usually originates directly from the aortic arch but often arises with the brachiocephalic trunk. This is known as a “bovine arch” configuration. The CCA usually has no branches in its cervical region. Occasionally, however, it may give off the superior thyroid artery, vertebral artery, ascending pharyngeal artery, and occipital or inferior thyroid artery. At the carotid bifurcation, the CCA divides into the external carotid artery (ECA) and the internal carotid artery (ICA). The ICA usually has no branching vessels in the neck. The ECA, which supplies the facial musculature, has multiple branches in the neck. The ICA may demonstrate an ampullary region of mild dilation just beyond its origin.

CAROTID ULTRASOUND EXAMINATION Carotid artery ultrasound examinations are performed with the patient supine, the neck slightly extended, and the head turned away from the side being examined. Some operators prefer to perform the examination at the patient’s side, whereas others prefer to sit at the patient’s head. The examination sequence also varies

950   PART III  ■  Small Parts, Carotid Artery, and Peripheral Vessel Sonography

E

I

A

B

FIGURE 25-2.  Carotid sonographic anatomy. A, Transverse image of the left carotid bifurcation. The larger, more lateral vessel is the internal carotid artery (I); E, external carotid artery. B, Color Doppler shows normal flow separation (arrow) in the proximal internal carotid artery.

with operator preference. This sequence includes the gray-scale examination, Doppler spectral analysis, and color Doppler blood flow interrogations. Power Doppler sonography may or may not be employed. A 5 to 12– MHz transducer is used for gray-scale imaging and a 3 to 7–MHz transducer for Doppler sonography; the choice depends on the patient’s body habitus and technical characteristics of the ultrasound machine. Color Doppler flow imaging and power Doppler imaging may be performed with 5 to 10–MHz transducers. In cases of critical stenosis, the Doppler parameters should be optimized to detect extremely slow flow. Gray-scale sonographic examination begins in the transverse projection. Scans are obtained along the entire course of the cervical carotid artery, from the supraclavicular notch cephalad to the angle of the mandible (Fig. 25-2). Inferior angulation of the transducer in the supraclavicular area images the CCA origin. The left CCA origin is deeper and more difficult to image consistently than the right. The carotid bulb is identified as a mild widening of the CCA near the bifurcation. Transverse views of the carotid bifurcation establish the orientation of the external and internal carotid arteries and help define the optimal longitudinal plane in which to perform Doppler spectral analysis. When the transverse ultrasound images demonstrate occlusive atherosclerotic disease, the percentage of “diameter stenosis” or “area stenosis” can be calculated directly using electronic calipers and software analytic algorithms available on most duplex equipment. After transverse imaging, longitudinal scans of the carotid artery are obtained. The examination plane necessary for optimal longitudinal scans is determined by the course of the vessels demonstrated on the transverse study. In some patients the optimal longitudinal orientation will be nearly coronal, whereas in others it will be almost sagittal. In most cases the optimal longitudinal scan plane will be oblique, between sagittal and coronal.

C

E

I

FIGURE 25-3.  Carotid bifurcation. Longitudinal image demonstrates common carotid artery (C); external carotid artery (E); and large, posterior internal carotid artery (I).

In approximately 60% of patients, both vessels above the carotid bifurcation and the CCA can be imaged in the same plane (Fig. 25-3); in the remainder, only a single vessel will be imaged in the same plane as the CCA. Images are obtained to display the relationship of both branches of the carotid bifurcation to the visualized plaque disease, and the cephalocaudal extent of the plaque is measured. Several anatomic features differentiate the ICA from the ECA. In about 95% of patients, the ICA is posterior and lateral to the ECA. This may vary considerably, however,10 and the ICA may be medial to the ECA in 3% to 9% of people. The ICA frequently has an ampullary region of dilation just beyond its origin and is usually larger than the ECA. One reliable distinguishing feature of the ECA is its branching vessels (Fig. 25-4, A). The superior thyroid artery is often seen as the first branch of the ECA after the bifurcation of the CCA.

Chapter 25  ■  The Extracranial Cerebral Vessels   951

A

B

FIGURE 25-4.  Normal external carotid artery (ECA). A, Color Doppler ultrasound of bifurcation demonstrates two small arteries originating from the ECA. B, ECA spectral Doppler shows the anticipated serrated (sawtooth) flow disturbance from the temporal artery tap (TT).

Occasionally, an aberrant superior thyroid artery branch will arise from the distal CCA. The ICA usually has no branches in the neck, although rarely the ICA gives rise to the ascending pharyngeal, occipital, facial, laryngeal, or meningeal arteries. In some patients, a considerable amount of the ICA will be visible, but in others, only the immediate origin of the vessel will be accessible. Very rarely, the bifurcation may not be visible at all.19 Rarely, the ICA may be hypoplastic or congenitally absent.20 A useful method to identify the ECA is the tapping of the superficial temporal artery in the preauricular area, the temporal tap (TT). The pulsations are transmitted back to the ECA, where they cause a sawtooth appearance on the spectral waveform (Fig. 25-4, B). Although the tap helps identify the ECA, this tap deflection may be transmitted into the CCA and even the ICA in certain rare situations.

CAROTID ULTRASOUND INTERPRETATION Each facet of the carotid sonographic examination is valuable in the final determination of the presence and extent of disease. In most cases the gray-scale, color Doppler, and power Doppler sonographic images and assessments will agree. However, when there are discrepancies between Doppler ultrasound and information, every attempt should be made to discover the source of the disagreement. The more closely the image and Doppler findings correlate, the higher the degree of confidence in the diagnosis. Generally, gray-scale and color

or power Doppler images better demonstrate and quantify low-grade stenoses, whereas high-grade occlusive disease is more accurately defined by Doppler spectral analysis. For plaque characterization, assessment must be made in gray scale only, without color or power Doppler ultrasound.

Visual Inspection of Gray-Scale Images Vessel Wall Thickness and Intima-Media Thickening Longitudinal views of the layers of the normal carotid wall demonstrate two nearly parallel echogenic lines, separated by a hypoechoic to anechoic region (Fig. 25-5). The first echo, bordering the vessel lumen, represents the lumen-intima interface; the second echo is caused by the media-adventitia interface. The media is the anechoic/ hypoechoic zone between the echogenic lines. The distance between these lines represents the combined thickness of the intima and media (I-M complex). The far wall of the CCA is measured. Many consider measurement of intima-media thickness (IMT) to be a surrogate marker for atherosclerotic disease in the whole arterial system, not only the cerebrovascular system. Some believe that thickening of the I-M complex greater than 0.8 mm is abnormal and may represent the earliest changes of atherosclerotic disease. However, because thickness of the I-M increases with age, absolute measurements of IMT for any given person may not

952   PART III  ■  Small Parts, Carotid Artery, and Peripheral Vessel Sonography

be a reliable indicator of atherosclerotic risk factors21 (Fig. 25-6). Numerous studies support the relationship between IMT and increased risk for myocardial infarction or stroke in asymptomatic patient populations.14,22-29 IMT may be superior to the coronary artery calcification score for identifying patients at high risk for these cardiovascular events.14 Assessment of IMT has been advocated as a means of assessing effectiveness of medical interventions to reduce the progression of I-M thickening or even reverse carotid wall thickening. Whether these

FIGURE 25-5.  Normal intima-media (I-M) complex of common carotid artery. The I-M complex (arrows) is seen in a left common carotid artery.

A

measurements have validity for assessment of an individual patient versus large groups of patients remains controversial. Studies demonstrating the accuracy of interobserver variability, reproducibility, and precision are needed before IMT assessment can be advocated for individual patient management.

Plaque Characterization Atheromatous carotid plaques should be carefully evaluated to determine plaque extent, location, surface contour, and texture, as well as to assess luminal stenosis.30 The plaque should be scanned and evaluated in both the sagittal and the transverse projections.31 The most common cause of TIAs is embolism, not flowlimiting stenosis; less than half of patients with documented TIA have hemodynamically significant stenosis. It is important to identify low-grade atherosclerotic lesions that may contain hemorrhage or ulceration, which can serve as a nidus for emboli that cause both TIAs and stroke.1 Polak et al.32 showed that plaque is an independent risk factor for developing a stroke.32 Of patients with hemispheric stroke symptoms, 50% to 70% demonstrate hemorrhagic or ulcerated plaque. Plaque analysis of CEA specimens has implicated intraplaque hemorrhage as an important factor in the development of neurologic symptoms.33-39 However, the relationship between sonographic plaque morphology and the onset of symptoms is controversial. Plaque texture is generally classified as homogeneous or heterogeneous.* The accurate evaluation of plaque can only be made with gray-scale ultrasound, without the use of color or power Doppler. The plaque must be evaluated in both sagittal and transverse planes.31 *References 9, 24, 27, 30, 31, 33-35, 40-44.

B

FIGURE 25-6.  Abnormal intima-media complex of common carotid artery (CCA). A, Early I-M hyperplasia with loss of the hypoechoic component of the I-M complex and thickening (arrows). B, Thickening of the I-M complex with hyperplasia (arrows).

Chapter 25  ■  The Extracranial Cerebral Vessels   953

A

B

C

D

FIGURE 25-7.  Homogeneous plaque. A, Sagittal, and B, transverse, images show homogeneous plaque in left common carotid artery (type 4). Note the uniform echo texture. C, Sagittal, and D, transverse, images show homogeneous plaque in proximal left internal carotid artery (type 3). Note the focal hypoechoic area within the plaque, estimated at less than 50% of plaque volume.

Homogeneous plaque has a generally uniform echo pattern and a smooth surface (Fig. 25-7). Sonolucent areas may be seen, but the amount of sonolucency is less than 50% of the plaque volume. The uniform acoustic texture corresponds pathologically to dense fibrous connective tissue. Calcified plaque produces posterior acoustic shadowing and is common in asymptomatic individuals (Fig. 25-8). Heterogeneous plaque has a more complex echo pattern and contains one or more focal sonolucent areas corresponding to more than 50% of the plaque volume (Fig. 25-9). Heterogeneous plaque is characterized pathologically by containing intraplaque hemorrhage and deposits of lipid, cholesterol, and proteinaceous material.10,42 Homogeneous plaque is identified much more often than heterogeneous plaque, occurring in 80% to 85% of patients examined.32 Sonography accurately determines the presence or absence of

intraplaque hemorrhage (sensitivity, 90%-94%; specificity, 75%-88%).33,39,42,45-47 Some sources suggest classifying plaque according to four types. Plaque types 1 and 2, similar to heterogeneous plaque and much more likely to be associated

ULTRASOUND TYPES OF PLAQUE MORPHOLOGY Type 1: Predominantly echolucent, with a thin echogenic cap Type 2: Substantially echolucent with small areas of echogenicity (>50% sonolucent) Type 3: Predominantly echogenic with small areas of echolucency (95%), the velocity measurements may actually decrease, and the waveform becomes dampened58,66 (Fig. 25-20, D and E ). In these cases, correlation with color or power Doppler imaging is essential to diagnose correctly the severity of the stenoses. Velocity increases are focal and most pronounced in and immediately distal to a stenosis, emphasizing the importance of sampling directly in these regions. As one moves further distal from a stenosis, flow begins to reconstitute and assume a more normal pattern, provided a tandem lesion does not exist distal to the initial site of stenosis. Spectral broadening results in the jets of high-velocity flow associated with carotid stenosis; however, correlation with gray-scale and color Doppler images can define other causes of spectral broadening. An awareness of normal flow spectra combined with appropriate Doppler techniques can obviate many potential diagnostic pitfalls. The degree of carotid stenosis that is considered clinically significant in the symptomatic or asymptomatic patient is in evolution. Initially, it was thought that lesions causing 50% diameter stenosis were significant; this perception changed as more information was gathered from two large clinical trials. As noted earlier, NASCET demonstrated that CEA was more beneficial than medical therapy in symptomatic patients with 70% to 99% ICA stenosis.4 ECST demonstrated a CEA benefit when the degree of stenosis was greater than 60%.5 Interestingly, the method used to grade stenoses in the ECST study was significantly different than that

used in the NASCET trials. The NASCET trials compared the severity of the ICA stenosis on arteriogram with the residual lumen of a presumably more normal distal ICA. The ECST methodology entailed assessment of the severity of stenosis with a “guesstimation” of the lumen of the carotid artery at the level of the stenosis. The ECST assessment is more comparable to ultrasound’s visible assessment of the degree of narrowing, whereas velocity tables currently in use have been derived to correspond to the NASCET angiographic determinations for stenosis. The ECST method for grading carotid artery stenosis tends to give a more severe assessment of narrowing than the NASCET technique (Fig. 25-21). The initial NASCET trials retrospectively compared velocity data obtained on the Doppler examination with angiographic measurements of stenosis. No standardized ultrasound protocol was employed by the numerous centers involved in the trials. Despite the lack of uniformity, moderate sensitivity and specificity ranging from 65% to 77% were obtained for grading ICA stenoses using Doppler velocities. If ultrasound technique is standardized and criteria are validated in a given laboratory, peak systolic velocity (PSV) and peak systolic ratios have proved to be an accurate method for determining carotid stenosis.67 The ECST group compared three different angiographic measurement techniques: the NASCET, the ECST, and a technique comparing distal CCA measurements with those of ICA stenosis. Researchers concluded that the ECST and NASCET techniques were similar in their prognostic value, whereas the CCA/ stenosis measurement was the most reproducible of the three techniques. They also concluded that the CCA method, although reproducible, would be invalidated by the presence of CCA disease.68 Virtually all investigators advocate using the NASCET angiographic measurement technique. The results of these trials, as well as the more recent ACAS and moderate NASCET studies, have generated reappraisals of the Doppler velocity criteria that most accurately define 70% or greater stenosis and, more recently, greater than 50% diameter stenoses.69 Attempts have been made to determine the Doppler parameters or combination of parameters that most reliably identify a certain-diameter stenosis. Most sources agree that the best parameter is the PSV of the ICA in the region of a stenosis.66 Using multiple parameters can improve diagnostic confidence, particularly when combined with color and power Doppler imaging. The degree of stenosis is best assessed using the grayscale and pulsed Doppler parameters, including ICA PSV, ICA end diastolic velocity (EDV), CCA PSV, CCA EDV, peak systolic ICA/CCA ratio (SVR), and peak end diastolic ICA/CCA ratio (EDR).66,67,70 Peak systolic velocity has proved accurate for quantifying high-grade stenoses.57,67 The relationship of PSV to the degree of luminal narrowing is well defined and easily measured.71,72 Although Doppler velocities have proved

Chapter 25  ■  The Extracranial Cerebral Vessels   963

B

A

C

E

D

FIGURE 25-20.  Internal carotid artery (ICA) stenosis. A, ICA stenosis of 50% to 69% diameter shows a peak systolic velocity (PSV) of 129 cm/sec. B, Right ICA demonstrates a visible high-grade stenosis on color Doppler with end diastolic velocities (EDVs) of greater than 288 cm/sec and PSVs that alias at greater than 400 cm/sec. This is consistent with a very high-grade stenosis. C, Left carotid bulb seen in longitudinal projection with color Doppler demonstrates a high-grade narrowing and spectral broadening with an approximately 400 cm/sec velocity in peak systole and 150 cm/sec in end diastole, consistent with an 80% to 95% stenosis. D, Velocity obtained in the ICA demonstrates low velocities. The PSV is 78 cm/sec, EDV is 22 cm/sec, systolic velocity ratio is 1.7, and diastolic velocity ratio is 2.8. E, The color-flow Doppler image demonstrates a markedly narrowed vessel. The degree of stenosis correlates with 95% to 99%.

964   PART III  ■  Small Parts, Carotid Artery, and Peripheral Vessel Sonography ICA C B

ECA

A

Measurement Methodology ECST =

B–A

x 100

B NASCAT = 1– ACAS

A

x 100

C

CCA

FIGURE 25-21.  Comparative measurement methodology. Different methodologies for grading internal carotid artery stenoses, from North American Symptomatic Carotid Endarterectomy Trial (NASCET), Asymptomatic Carotid Atherosclerosis Study (ACAS), and European Carotid Surgery Trial (ECST).

reliable for defining 70% or greater stenosis, Grant et al.67 showed less favorable results for substenosis classification between 50% and 69% using PSV and ICA/ CCA PSV ratios. In our experience, however, using all four parameters and determining a correct category for the degree of stenosis is the most efficacious way to ensure accuracy. Agreement for all four parameters for a clinical situation is most common. When there is an outlying parameter, further assessment and careful attention to technique and detail are required. EDV and EDR are particularly useful in distinguishing between high grades of stenosis. Additionally, correlating the visual estimation of the degree of stenosis and the velocity numbers will help in correctly grading stenosis, particularly when the degree of stenosis is “near occlusion” (Figs. 25-22 and 25-23; see also Fig. 25-20, D and E). On rare occasions, alternate imaging methodologies (e.g., MRA, CT angiography) may need to be recommended. No criteria for grading external carotid artery stenoses have been established. A good general rule is that if the ECA velocities do not exceed 200 cm/sec, no significant stenosis is present. However, we usually rely on a visible assessment of the degree of narrowing associated with velocity changes. Occlusive plaque involving the ECA is less common than in the ICA and is rarely clinically significant. Similarly, velocity criteria used to grade common carotid artery stenoses have not been well established. However, if one is able to visualize 2 cm proximal and 2 cm distal to a visible CCA stenosis, a PSV ratio obtained 2 cm proximal to the stenosis (vs. in region of greatest visible stenosis) can be used to grade

the “percent diameter stenosis” in a manner similar to that used in peripheral artery studies. A doubling of the PSV across a lesion would correspond to at least a 50% diameter stenosis, and a velocity ratio in excess of 3.5 corresponds to a greater than 75% stenosis. Although duplex ultrasound remains an accurate method of quantifying ICA stenoses, the use of color and power Doppler sonography has significantly improved diagnostic confidence and reproducibility.73 One persistent problem with duplex Doppler with gray-scale ultrasound evaluation of the carotids is that different institutions use PSVs ranging from 130 cm/ sec74 to 325 cm/sec69 to diagnose greater than 70% ICA stenosis. Factors creating these discrepancies include technique and equipment.75 This wide range of PSVs reinforces the need for individual ultrasound laboratories to determine which Doppler parameters are most reliable in their own institution.75 Correlation of the velocity ranges obtained by ultrasound with angiographic and surgical results is necessary to achieve accurate, reproducible examinations in a particular ultrasound laboratory.76 The Society of Radiologists in Ultrasound, representing multiple medical and surgical specialties, held a consensus conference in 2002 to consider carotid Doppler ultrasound.77 In addition to guidelines for performing and interpreting carotid ultrasound examinations, panelists devised a set of criteria widely applicable among vascular laboratories (Table 25-1).77 Although the conference did not recommend all established laboratories with internally validated velocity charts alter their practices, they suggested physicians establishing new laboratories consider using the consensus criteria; those with preexisting charts might consider comparing in-house criteria with those provided by the consensus conference. Velocity criteria corresponding to specific degrees of vascular stenosis are listed in the tables. Our institution uses Table 25-2, which has a category for 80% to 95% stenoses; our surgeons are more inclined to consider surgery for patients with asymptomatic stenoses greater than 80% than for those with less severe stenoses.30,78 The ICA values should be obtained at or just distal to the point of maximum visible stenosis and at the point of greatest color Doppler spectral abnormality. Values from the CCA should be obtained 2 cm proximal to the widening in the region of the carotid bulb. Because velocities normally decrease from proximal to distal in the CCA and increase from proximal to distal in the ICA, it is important that standardized levels be used routinely for obtaining the ICA/CCA velocity ratio.

Color Doppler Ultrasound Color Doppler ultrasound displays blood flow information in real time over the entire image or a selected area. Stationary soft tissue structures, which lack a detectable phase or frequency shift, are assigned an amplitude value

Chapter 25  ■  The Extracranial Cerebral Vessels   965

A

D

C

B

E

FIGURE 25-22.  Abnormal high-resistance waveforms. High-resistance waveforms: A, common carotid artery; B, proximal internal carotid artery (ICA); and D, distal ICA. Color flow Doppler imaging of the carotid bulb in transverse (D) and sagittal (E) projections demonstrates a significantly narrowed ICA. These findings are consistent with a greater than 95% stenosis of the ICA and a distal tandem stenosis of the intracranial carotid artery.

966   PART III  ■  Small Parts, Carotid Artery, and Peripheral Vessel Sonography

A

B

C

D

E FIGURE 25-23.  Near occlusion (95-99% stenosis) with homogeneous plaque. A, Transverse, and B, sagittal, grayscale images of the left internal carotid artery (ICA) demonstrate homogeneous (type 3) plaque. C, Transverse, and D, sagittal, power Doppler images demonstrate extremely narrowed residual lumen. E, Velocity measurements for the ICA were peak systolic velocity, 51 cm/ sec; acoustic Doppler velocity, 19 cm/sec; systolic velocity ratio, 51/64 = 0.8; diastolic velocity ratio, 19/12 = 1.5. The combination of visual images and Doppler spectral analysis findings indicate a 95% to 99% stenosis.

Chapter 25  ■  The Extracranial Cerebral Vessels   967

TABLE 25-1.  DIAGNOSTIC CRITERIA FOR CAROTID ULTRASOUND EXAMINATIONS

Normal 4 >2.15 ≥2.7 ≥4.15 >2 >4 2.45 4.3

Modified from Fleming et al.126 and Chahwan et al.131 PSV, Peak systolic velocity; EDV, end diastolic velocity; ICA/CCA, internal/common carotids.

>125

Chapter 25  ■  The Extracranial Cerebral Vessels   981

A

B

C FIGURE 25-39.  Fibromuscular dysplasia. A, Longitudinal color Doppler image of the middle to distal portion of the internal carotid artery (ICA) shows velocity elevation and significant stenosis. B, Same patient’s proximal portion of the ICA shows no stenosis. C, Angiogram demonstrates typical appearance of fibromuscular dysplasia in the mid-ICA and distal ICA. Note the beaded appearance resulting from focal bands (arrow) of thickened tissue that narrow the lumen.

characteristic “string of beads” appearance has been described on angiography. Only a few reports describe sonographic features of FMD.137,138 Many patients with FMD demonstrate nonspecific or no obvious abnormalities on ultrasound. FMD may be asymptomatic or can result in carotid dissection or subsequent thromboembolic events (Fig. 25-39). Arteritis resulting from autoimmune processes (e.g., Takayasu’s arteritis, temporal arteritis) or radiation changes can produce diffuse con-

centric thickening of carotid walls, which most frequently involves the CCA139 (Fig. 25-40). Cervical trauma can produce carotid dissections or aneurysms. Carotid artery dissection results from a tear in the intima, allowing blood to dissect into the wall of the artery, which produces a false lumen. The false lumen may be blind ended or may reenter the true lumen. The false lumen may occlude or narrow the true lumen, producing symptoms similar to carotid

982   PART III  ■  Small Parts, Carotid Artery, and Peripheral Vessel Sonography

A

B

C FIGURE 25-40.  Long-segment stenosis of common carotid artery (CCA) caused by Takayasu’s arteritis. A, Power Doppler image of left CCA shows long-segment concentric narrowing caused by greatly thickened walls of the artery. B, Power Doppler image of right CCA in same patient demonstrates similar concentric narrowing (arrows). C, Right spectral Doppler waveform shows a mildly tardus-parvus waveform.

plaque disease. Dissections may arise spontaneously or secondary to trauma or to intrinsic disease with elastic tissue degeneration (e.g., Marfan’s syndrome) or may be related to atherosclerotic plaque disease.15 The ultrasound examination of a carotid dissection may reveal a mobile or fixed echogenic intimal flap, with or without thrombus formation. Frequently, there is a striking image/Doppler mismatch with a paucity of gray-scale abnormalities seen in association with marked flow abnormalities (Fig. 25-41). Color or power Doppler ultrasound may readily clarify the source of this mismatch by demonstrating abrupt tapering of the patent, filled lumen to the point of an ICA occlusion, analogous to angiographic findings. Although the ICA is frequently occluded, demonstrating absent flow with a high-resistance waveform in the proxi-

mal ipsilateral CCA, flow in the ICA may demonstrate high velocities associated with luminal narrowing secondary to hemorrhage and a thrombus in the area of the false lumen. Accordingly, flow velocity waveforms in the CCA may be normal or may demonstrate extremely damped, high-resistance waveforms. MRA, another noninvasive imaging test, readily demonstrates mural hematoma that confirms the diagnosis of ICA dissection. Although angiography is frequently used initially to diagnose a dissection, ultrasound can be used to follow patients to assess the therapeutic response to anticoagulation. Repeat sonographic evaluation of patients with ICA dissection after anticoagulation therapy reveals recanalization of the artery in as many as 70% of cases.140-142 It is important to consider the diagnosis of dissection as a cause of neurologic symptoms, particu-

Chapter 25  ■  The Extracranial Cerebral Vessels   983

A

C

B

D

E

FIGURE 25-41.  Carotid artery dissection. A, Abnormal high-resistance waveforms (arrow) at the origin of the right internal carotid artery (ICA) with no evidence of flow distal to this point (curved arrow). B, Gray-scale evaluation of the vessel in the area of the occlusion demonstrates only a small, linear echogenic structure (arrow) without evidence of significant atherosclerotic narrowing. C, Subsequent angiogram demonstrates the characteristic tapering to the point of occlusion (arrow) associated with carotid artery dissection and thrombotic occlusion. D, Transverse, and E, longitudinal, images of another patient show an intimal flap (arrow) in an external carotid artery; I, internal carotid artery.

INTERNAL CAROTID ARTERY DISSECTION: SONOGRAPHIC FINDINGS Internal Carotid Artery

Absent flow or occlusion Echogenic intimal flap, with or without thrombus Hypoechoic thrombus, with or without luminal narrowing Normal appearance

Common Carotid Artery High-resistance waveform Damped flow Normal appearance

larly when the clinical presentation, age, and patient history are atypical for that of atherosclerotic disease or hemorrhagic stroke. The most common CCA aneurysm occurs in the region of the carotid bifurcation. These aneurysms may result from atherosclerosis, infection, trauma, surgery, or infectious etiology, such as syphilis. The normal CCA usually measures no more than 1 cm in diameter. Carotid body tumors, one of several paragangliomas that involve the head and neck, are usually benign, wellencapsulated masses located at the carotid bifurcation. These tumors may be bilateral, particularly in the familial variant, and are very vascular, often producing an audible bruit. Some of these tumors produce catecholamines, leading to sudden changes in blood pressure during or after surgery. Color Doppler ultrasound demonstrates an extremely vascular soft tissue mass at the carotid bifurcation (Fig. 25-42). Color Doppler

984   PART III  ■  Small Parts, Carotid Artery, and Peripheral Vessel Sonography

ECA

ICA

A

B

FIGURE 25-42.  Carotid body tumor. A, Transverse image of the carotid bifurcation shows a mass (arrows) splaying the internal carotid artery (ICA) and external carotid artery (ECA). B, Pulsed Doppler traces of the carotid body tumor show typical arteriovenous shunt (low-resistance) waveform.

ultrasound can also be used to monitor embolization or surgical resection of carotid body tumors. A classic nonmass is the ectatic innominate/proximal CCA, frequently occurring as a pulsatile supraclavicular mass in older women. The request to rule out a carotid aneurysm almost invariably shows the classic normal features of these tortuous vessels (Fig. 25-43). Extravascular masses (e.g., lymph nodes, hematomas, abscesses) that compress or displace the carotid arteries can be readily distinguished from primary vascular masses, such as aneurysms or pseudoaneurysms (Fig. 25-44). Posttraumatic pseudoaneurysms can usually be distinguished from a true carotid aneurysm by demonstrating the characteristic to-and-fro waveforms in the neck of the pseudoaneurysm, as well as the internal variability (yin-yang) characteristic of a pseudoaneurysm (Fig. 25-45).

I

TRANSCRANIAL DOPPLER SONOGRAPHY In transcranial Doppler (TCD) ultrasound, a low-frequency 2-MHz transducer is used to evaluate blood flow within the intracranial carotid and vertebrobasilar system and the circle of Willis. Access is achieved through the orbits, foramen magnum, or most often the region of temporal calvarial thinning (transtemporal window).143 However, many patients (up to 55% in one series144) may not have access to an interpretable TCD examination. Women, particularly African Americans, have a thick temporal bone through which it is difficult to

FIGURE 25-43.  Ectatic common carotid artery (CCA). Color Doppler image shows ectatic proximal CCA arising from the innominate artery (I) and responsible for a pulsatile right supraclavicular mass.

insonate the basal cerebral arteries.144,145 This difficulty limits the feasibility of TCD imaging as a routine part of the noninvasive cerebrovascular workup.144 By using spectral analysis, various parameters are determined, including mean velocity, PSV, EDV, and

Chapter 25  ■  The Extracranial Cerebral Vessels   985

P

C

E

I

FIGURE 25-44.  Pathologic lymph node near carotid bifurcation. Power Doppler image shows a malignant lymph node (arrow) lateral to the carotid bifurcation.

the pulsatility and resistive indices of the blood vessels. Color or power Doppler ultrasound can improve velocity determination by providing better angle theta determination and localizing the course of vessels.143 TCD applications include (1) evaluation of intracranial stenoses and collateral circulation, (2) detection and follow-up of vasoconstriction from subarachnoid hemorrhage, (3) determination of brain death, (4) evaluation of patients with sickle cell disease, and (5) identification of arteriovenous malformation.140-143,146 TCD is most reliable in diagnosing stenoses of the middle cerebral artery (MCA), with sensitivities as high as 91% reported. TCD is less reliable for detecting stenoses of the intracranial vertebrobasilar system, anterior and posterior cerebral arteries, and terminal ICA. However, TCD is helpful in assessing vertebral artery patency and flow direction when no flow is detected in the extracranial vertebral artery (Fig. 25-46). Diagnosis of an intracranial stenosis is based on an increase in the mean velocity of blood flow in the affected vessel compared to that of the contralateral vessel at the same location.144,145 Advantages of TCD ultrasound also include its availability to monitor patients in the operating room or angiographic suite for potential cerebrovascular com­ plications.145 Intraoperative TCD monitoring can be performed with the transducer strapped over the transtemporal window, allowing evaluation of blood flow in

FIGURE 25-45.  Pseudoaneurysm of the common carotid artery (CCA). Transverse image of the left distal CCA (C) demonstrates a characteristic to-and-fro waveform in the neck of the large pseudoaneurysm (P), which resulted from an attempted central venous line placement.

the MCA during CEA. The adequacy of cerebral perfusion can be assessed while the carotid artery is clamped.145,147 TCD is also capable of detecting intraoperative microembolization (“HITS”), which produces high-amplitude spikes on the Doppler spectrum.144,148-150 The technique can be used for the serial evaluation of vasospasms. This diagnosis is usually based on serial examinations of the relative increase in blood flow velocity and resistive index changes resulting from a decrease in the lumen of the vessel caused by vasospasms.145

VERTEBRAL ARTERY The vertebral arteries supply the majority of the posterior brain circulation. Through the circle of Willis, the vertebral arteries also provide collateral circulation to other portions of the brain in patients with carotid occlusive disease. Evaluation of the extracranial vertebral artery seems a natural extension of carotid duplex and color Doppler imaging.151,152 Historically, however, these arteries have not been studied as intensively as the carotids. Symptoms of vertebrobasilar insufficiency

986   PART III  ■  Small Parts, Carotid Artery, and Peripheral Vessel Sonography

A

B

FIGURE 25-46.  Transcranial Doppler imaging. A, Transcranial duplex scan of the posterior fossa in a patient with an incomplete left subclavian steal syndrome demonstrates retrograde systolic flow (arrow) and antegrade diastolic flow (curved arrow). The scan is obtained in a transverse projection from the region of the foramen magnum (open arrowhead). B, Color Doppler image obtained in the same patient demonstrates that there is retrograde flow not only within the left vertebral artery, but within the basilar artery (arrow) as well.

also tend to be rather vague and poorly defined compared with symptoms referable to the carotid circulation. It is often difficult to make an association confidently between a lesion and symptoms. Furthermore, interest in surgical correction of vertebral lesions has been limited. The anatomic variability, small size, deep course, and limited visualization resulting from overlying transverse processes make the vertebral artery more difficult to examine accurately with ultrasound.151,153,154 The clinical utility of vertebral artery duplex scanning remains under investigation. Its role in diagnosing subclavian steal and presteal phenomena is well established.155,156 Less clear-cut is the use of vertebral duplex scanning in evaluating vertebral artery stenosis, dissection, or aneurysm.

Anatomy The vertebral artery is usually the first branch off the subclavian artery (Fig. 25-47). However, variation in the origin of the vertebral arteries is common. In 6% to 8% of people the left vertebral artery arises directly from the aortic arch proximal to the left subclavian artery. In 90% the proximal vertebral artery ascends superomedially, passing anterior to the transverse process of the seventh cervical vertebra (C7), and enters the transverse foramen at the C6 level. The remainder of vertebral arteries enters into the transverse foramen at the C5 or C7 level and, rarely, at the C4 level. The size of vertebral arteries is variable, with the left larger than the right in 42%, the two vertebral arteries equal in size in 26%, and the right

B

C

S

FIGURE 25-47.  Vertebral artery course. Lateral diagram of vertebral artery (arrow) shows its course through the cervical spine transverse foramina (arrowheads) en route to joining the contralateral vertebral artery to form the basilar artery (B); C, carotid artery; S, subclavian artery.

larger than the left in 32% of cases.157 One vertebral artery may even be congenitally absent. Usually, the vertebral arteries join at their confluence to form the basilar artery. Rarely, the vertebral artery may terminate in a posterior inferior cerebellar artery.

Chapter 25  ■  The Extracranial Cerebral Vessels   987

V S A

FIGURE 25-48.  Normal vertebral artery and vein. Longitudinal color Doppler image shows a normal vertebral artery (A) and vein (V) running between the transverse processes of the second to sixth cervical vertebrae (C2-C6), which are identified by their periodic acoustic shadowing (S).

Sonographic Technique and Normal Examination Vertebral artery visualization with Doppler flow analysis can be obtained in 92% to 98% of vessels151,158 (Fig. 25-48). Color Doppler facilitates the rapid detection of vertebral arteries but does not significantly improve this detection rate.154 Vertebral artery duplex examinations are performed by first locating the CCA in the longitudinal plane. The direction of flow in the CCA and jugular vein is determined. A gradual sweep of the transducer laterally demonstrates the vertebral artery and vein running between the transverse processes of C2 to C6, which are identified by their periodic acoustic shadowing. Transverse scanning with color Doppler allows the examiner to visualize the carotid artery and jugular vein at the same time and use them as references to determine the direction of flow in the vertebral artery.153,155 Angling the transducer caudad allows visualization of the vertebral artery origin in 60% to 70% of the arteries, in 80% on the right-hand side, and in 50% on the left. This discrepancy may relate to the left vertebral artery origin being deeper and arising directly from the aortic arch in 6% to 8% of cases.153,159 The presence and direction of flow should be established. Visible plaque disease should be assessed. The vertebral artery supplies blood to the brain and usually has a low-resistance flow pattern similar to that of the CCA, with continuous flow in systole and diastole; however, wide variability in waveform shape has been noted in angiographically normal vessels.160 Because the

FIGURE 25-49.  Normal vertebral artery waveform. Normal low-resistance waveform of the vertebral artery with filling of the spectral window.

vessel is small, flow tends to demonstrate a broader spectrum. The clear spectral window seen in the normal carotid system is often filled in the vertebral artery61 (Fig. 25-49). The vertebral vein (often a plexus of veins) runs parallel and adjacent to the vertebral artery. Care must be taken not to mistake its flow for that of the adjacent artery, particularly if the venous flow is pulsatile. Comparison with jugular venous flow during respiration should readily distinguish between vertebral artery and vein. At times, the ascending cervical branch of the thyrocervical trunk can be mistaken for the vertebral artery. This can be avoided by looking for landmark transverse processes that accompany the vertebral artery and by paying careful attention to the waveform of the visualized vessel. The ascending cervical branch has a highimpedance waveform pattern similar to that of the ECA.155 Transcranial Doppler sonographic examination of the vertebrobasilar artery system can be performed as an adjunct to the extracranial evaluation. The examination is conducted with a 2-MHz transducer with the patient sitting, using a suboccipital midline nuchal approach, or with the patient supine, using a retromastoidal approach. Color or power Doppler facilitates TCD imaging of the vertebrobasilar system.161

Subclavian Steal The subclavian steal phenomenon (syndrome) occurs when there is high-grade stenosis or occlusion of the proximal subclavian or innominate arteries with patent vertebral arteries bilaterally. The artery of the ischemic limb “steals” blood from the vertebrobasilar circulation through retrograde vertebral artery flow, which may

988   PART III  ■  Small Parts, Carotid Artery, and Peripheral Vessel Sonography

ABNORMAL VERTEBRAL ARTERY WAVEFORMS

W

Complete Subclavian Steal

Reversal of flow within vertebral artery ipsilateral to stenotic or occluded subclavian or innominate artery.

Incomplete or Partial Subclavian Steal

Transient reversal of vertebral artery flow during systole. May be converted into a complete steal using provocative maneuvers. Suggests stenotic, not occlusive, lesion. R

Presteal Phenomenon

L

“Bunny” waveform: systolic deceleration less than diastolic flow. May be converted into partial steal by provocative maneuvers. Seen with proximal subclavian stenosis. S

FIGURE 25-50.  Hemodynamic pattern in subclavian steal syndrome. Proximal left subclavian artery occlusive lesion (arrowhead) decreases flow to the distal subclavian artery (S). This produces retrograde flow (large arrows) down the left vertebral artery (L) and stealing from the right vertebral artery (R) and other intracranial vessels through the circle of Willis (W).

result in symptoms of vertebrobasilar insufficiency (Fig. 25-50). Symptoms are usually most pronounced during exercise of the upper extremity but can be produced by changes in head position. However, there is often poor correlation between vertebrobasilar symptoms and the subclavian steal phenomenon. Usually, flow within the basilar artery is unaffected unless severe stenosis of the vertebral artery supplying the steal exists.161 Also, surgical or angioplastic restoration of blood flow may not result in relief of symptoms.162 The subclavian steal syndrome is most often caused by atherosclerotic disease, although traumatic, embolic, surgical, congenital, and neoplastic factors have also been implicated. Although the proximal subclavian stenosis or occlusion may be difficult to image, particularly on the left, the vertebral

Tardus-Parvus or Damped Waveform Seen with vertebral artery stenosis.

artery waveform abnormalities correlate with the severity of the subclavian disease. Doppler evaluation of the vertebral artery reveals four distinct abnormal waveforms that correlate with subclavian or vertebral artery pathology on angiography. These include the complete subclavian steal, partial or incomplete steal, presteal phenomenon, and tardus-parvus vertebral artery waveforms.160 In a complete subclavian steal, flow within the vertebral artery is completely reversed (Fig. 25-51). Incomplete steal or partial steal demonstrate transient reversal of vertebral flow during systole161,163 (Fig. 25-52). Incomplete steal suggests highgrade stenosis of the subclavian or innominate artery rather than occlusion. Provocative maneuvers, such as exercising the arm for 5 minutes or 5-minute inflation of a sphygmomanometer on the arm to induce rebound hyperemia on the side of the subclavian or innominate lesion, can enhance the sonographic findings and convert an incomplete steal to a complete steal.97,115 The presteal (“bunny”) waveform shows antegrade flow but with a striking deceleration of velocity in peak systole to a level less than EDV. This is seen in patients with proximal subclavian stenosis, which is usually less severe than in those with partial steal waveform.163 The bunny waveform can be converted into a partial steal or complete steal waveform by provocative maneuvers, such as the use of a blood pressure cuff (Fig. 25-53). A damped, tardus-parvus waveform can be seen in patients with high-grade proximal vertebral stenosis.156,163 With a subclavian steal, color Doppler may show two similarly color-encoded vessels between the transverse

Chapter 25  ■  The Extracranial Cerebral Vessels   989

A

B

FIGURE 25-51.  Vertebral artery flow. A, Subclavian steal causes reversed flow in vertebral artery. Complete vertebral artery flow reversal results from a right subclavian artery occlusion. Flow in this vertebral artery is toward the transducer. B, Slightly aberrant vertebral artery with color flow reversal.

Stenosis and Occlusion

FIGURE 25-52.  Incomplete subclavian steal. Flow in early systole is antegrade, flow in peak systole is retrograde, and flow in late systole and diastole (arrow) is again antegrade.

processes, representing the vertebral artery and vein.83 Transverse images of the vertebral artery with color Doppler show reversed flow compared with those of the CCA. A Doppler spectral waveform must be produced in all such cases to avoid mistaking flow reversal within an artery for flow in a pulsatile vertebral vein.83,155

Diagnosis of vertebral artery stenosis is more difficult than diagnosis of flow reversal. Most hemodynamically significant stenoses occur at the vertebral artery origin, situated deep in the upper thorax and seen in only 60% to 70% of patients.153,158,159 Even if the vertebral artery origin off the subclavian is visualized, optimal adjustments of the Doppler angle for accurate velocity measurements may be difficult because of the deep location and vessel tortuosity. No accurate reproducible criteria for evaluating vertebral artery stenosis exist. Because flow is normally turbulent within the vertebral artery, spectral broadening cannot be used as an indicator of stenosis. Velocity measurements are not reliable as criteria for stenosis because of the wide normal variation in vertebral artery diameter. Although velocities greater than 100 cm/ sec often indicate stenosis, they can occur in angiographically normal vessels. For example, high-flow velocity may be present in a vertebral artery that is serving as a major collateral pathway for cerebral circulation in cases of carotid occlusion21,119,164 (Fig. 25-54). Thus, only a focal increase in velocity of at least 50%, visible stenosis on gray-scale or color Doppler, or a striking tardus-parvus vertebral artery waveform is likely to indicate significant vertebral stenosis. The variability of resistive indices in normal and abnormal vertebral arteries precludes the use of this parameter as an indicator of vertebral disease.160

990   PART III  ■  Small Parts, Carotid Artery, and Peripheral Vessel Sonography

A

B

FIGURE 25-53.  Incomplete subclavian steal and provocative maneuver. A, Presteal left vertebral artery waveform. Flow decelerates in peak systole but does not reverse. B, After provocative maneuver, there is reversal of flow in peak systole in response to a decrease in peripheral arterial pressure.

FIGURE 25-54.  Increased flow velocity in vertebral artery. Pulsed Doppler spectral trace from a left vertebral artery demonstrates strikingly high velocities and disturbed flow (arrow). Although this degree of velocity elevation and flow disturbance could be associated with a focal stenosis, in this case there was increased velocity throughout the vertebral artery from bilateral internal carotid artery occlusion and increased collateral flow into the vertebral artery.

Chapter 25  ■  The Extracranial Cerebral Vessels   991

Diagnosis of vertebral artery occlusion is also difficult. Often, the inability to detect arterial flow results from a small or congenitally absent vertebral artery or a technically difficult examination. The differentiation of severe stenosis from occlusion is difficult for the same reasons. Extremely damped blood flow velocity in highgrade stenoses and a decreased number of RBCs traversing the area evaluated may result in a Doppler signal with amplitude too low to be detected.154 Power Doppler imaging may prove useful in this situation. Visualization of only a vertebral vein may indicate vertebral artery occlusion or congenital absence.

J

INTERNAL JUGULAR VEINS The internal jugular veins are the major vessels responsible for the return of venous blood from the brain. The most common clinical indication for duplex and color Doppler flow ultrasound of the internal jugular vein is the evaluation of suspected jugular venous thrombosis.165-172 Thrombus formation may be related to central venous catheter placement. Other indications include a diagnosis of jugular venous ectasia171-174 and guidance for internal jugular or subclavian vein cannulation,175-179 particularly in difficult situations where vascular anatomy is distorted.

Sonographic Technique The normal internal jugular vein is easily visualized. The vein is scanned with the neck extended and the head turned to the contralateral side. Longitudinal and transverse scans are obtained with light transducer pressure on the neck to avoid collapsing the vein. A coronal view from the supraclavicular fossa is used to image the lower segment of the internal jugular vein and the medial segment of the subclavian vein as they join to form the brachiocephalic vein. The jugular vein lies lateral and anterior to the CCA, lateral to the thyroid gland, and deep to the sternocleidomastoid muscle. The vessel has sharply echogenic walls and a hypoechoic or anechoic lumen. Normally, a valve can be visualized in its distal portion.168,176,180 The right internal jugular vein is usually larger than the left.175 Real-time ultrasound demonstrates venous pulsations related to right heart contractions, as well as changes in venous diameter that vary with changes in intrathoracic pressure. Doppler examination graphically depicts these flow patterns (Fig. 25-55). On inspiration, negative intrathoracic pressure causes flow toward the heart and the jugular veins to decrease in diameter. During expiration and during Valsalva maneuver, increased intrathoracic pressure causes a decrease in the blood return, and the veins enlarge, with minimal or no flow noted. The walls of the normal jugular vein collapse

FIGURE 25-55.  Normal jugular vein. Complex venous pulsations in a normal jugular vein (J) reflect the cycle of events in the right atrium.

completely when moderate transducer pressure is applied. Sudden patient sniffing reduces intrathoracic pressure, causing momentary collapse of the vein on real-time ultrasound, accompanied by a brief increase in venous flow toward the heart as shown by Doppler.167,169,171

Thrombosis Clinical features of jugular venous thrombosis (JVT) include a tender, poorly defined, nonspecific neck mass or swelling. The correct diagnosis may not be immediately obvious.168 Thrombosis of the internal jugular vein can be completely asymptomatic because of the deep position of the vein and the presence of abundant collateral circulation.171 This condition was previously diagnosed by venography, an invasive procedure prompted only by a high index of suspicion. With the introduction of noninvasive techniques, such as ultrasound, CT,181and MRA,182 JVT is being identified more frequently. Internal jugular thrombosis most often results from complications of central venous catheterization.166,170,171 Other causes include intravenous drug abuse, mediastinal tumor, hypercoagulable states, neck surgery, and local inflammation or adenopathy.168 Some cases are idiopathic or spontaneous.168 Possible complications of JVT include suppurative thrombophlebitis, clot propagation, and pulmonary embolism.168,172 Real-time examination reveals an enlarged, noncompressible vein, which may contain a visible echogenic intraluminal thrombus. An acute thrombus may be anechoic and indistinguishable from flowing blood; however, the characteristic lack of compressibility and absent Doppler or color Doppler flow in the region of a thrombus quickly lead to the correct diagnosis. In addition, there is visible loss of vein response to respiratory maneuvers and venous pulsation. Spectral and color

992   PART III  ■  Small Parts, Carotid Artery, and Peripheral Vessel Sonography

C C

A

B

C

D

E

F

FIGURE 25-56.  Internal jugular vein thrombosis: spectrum of appearances. A, Transverse image of an acute left internal jugular vein thrombus (arrow). The vein is distended and noncompressible. C, Common carotid artery. B, Longitudinal image of a different patient demonstrates a hypoechoic thrombus and no Doppler signal. C, Longitudinal color Doppler image shows a small amount of thrombus arising from the posterior wall of the internal jugular vein (IJV). D, Transverse image shows an echogenic thrombus, indicating chronic thrombus in IJV. E, Longitudinal image demonstrates a thrombus around jugular vein catheter. F, Longitudinal images show a thrombus arising from anterior wall. This thrombus probably results from previous catheter placement in this region.

Doppler interrogations reveal absent flow (Fig. 25-56). Collateral veins may be identified, particularly in cases of chronic internal jugular vein thrombosis. Central liquefaction or other heterogeneity of the thrombus also suggests chronicity. Chronic thrombi may be difficult to visualize because they tend to organize and are difficult to separate from echogenic perivascular fatty tissue.176 The absence of cardiorespiratory plasticity in a patent jugular or subclavian vein can indicate a more central, nonocclusive thrombus (Fig. 25-57). The confirmation of bilateral loss of venous pulsations strongly supports a more central thrombus, which can be documented by venography or MRA. A thrombus that is related to catheter insertion is often demonstrated at the tip of the catheter, although it may be seen anywhere along the course of the vein. The catheter can be visualized as two parallel echogenic lines separated by an anechoic region. Flow is not usually demonstrated in the catheter, even if the catheter itself is patent.

Sonography has proved to be a reliable means of diagnosing jugular and subclavian vein thrombosis and has the advantage over CT and MRI of being inexpensive, portable, and nonionizing and of requiring no intravenous contrast. Sonography has limited access and cannot image all portions of the jugular and subclavian veins, especially those located behind the mandible or below the clavicle, although knowledge of the full extent of a thrombus is not typically a critical factor in treatment planning.168,172 Serial sonographic examination to evaluate response to therapy after the initial assessment can be performed safely and inexpensively. Sonography can also document venous patency before vascular line placement, facilitating safer and more successful catheter insertion.

Acknowledgment Thanks to Rita Premo and Barbara Siede for their assistance with manuscript preparation.

Chapter 25  ■  The Extracranial Cerebral Vessels   993

A

B

C FIGURE 25-57.  Normal and abnormal waveforms. A, Brachiocephalic vein has normal cardiorespiratory change in the venous waveforms, implying a patent superior vena cava. B, Near-occlusive left central brachiocephalic vein stenosis caused by a prior central venous catheter in another patient. Pulsed Doppler waveform shows reversed nonpulsatile flow in the internal jugular vein (IJV). C, Left subclavian vein (SCV) shows centrally directed but monophasic flow toward an area of central collaterals (arrow) in patient with a malfunctioning left arteriovenous dialysis fistula.

References 1. Carroll BA. Carotid sonography. Radiology 1991;178:303313. 2. Executive Committee, Asymptomatic Carotid Atherosclerosis Study. Endarterectomy for asymptomatic carotid artery stenosis. JAMA 1995;273:1421-1428. 3. Fontenelle LJ, Simper SC, Hanson TL. Carotid duplex scan versus angiography in evaluation of carotid artery disease. Am Surg 1994;60:864-868. 4. North American Symptomatic Carotid Endarterectomy Trial Collaborators. Beneficial effect of carotid endarterectomy in symptom-

atic patients with high-grade carotid stenosis. N Engl J Med 1991;325:445-453. 5. European Carotid Surgery Trialists’ Collaborative Group. Interim results for symptomatic patients with severe (70-99%) or with mild (0-29%) carotid stenosis. MRC European Carotid Surgery Trial. Lancet 1991;337:1235-1243. 6. Barnett HJ, Taylor DW, Eliasziw M, et al. Benefit of carotid endarterectomy in patients with symptomatic moderate or severe stenosis. North American Symptomatic Carotid Endarterectomy Trial Collaborators. N Engl J Med 1998;339:1415-1425. 7. Bluth EI, Bertino RE, Grant EG, et al. ACR guideline for the performance of an ultrasound examination of the extracranial cerebrovascular system. Am Coll Radiol 2007:1037-1040.

994   PART III  ■  Small Parts, Carotid Artery, and Peripheral Vessel Sonography 8. Derdeyn CP, Powers WJ, Moran CJ, et al. Role of Doppler ultrasound in screening for carotid atherosclerotic disease. Radiology 1995;197:635-643. 9. Merritt CR, Bluth EI. The future of carotid sonography. AJR Am J Roentgenol 1992;158:37-39. 10. Taylor KJW. Clinical applications of carotid Doppler ultrasound. In: Taylor KJW, Burns PN, Wells PNT, editors. Clinical applications of Doppler ultrasound. New York: Raven Press; 1988. p. 120-161. 11. Bluth EI, Shyn PB, Sullivan MA, Merritt CR. Doppler color flow imaging of carotid artery dissection. J Ultrasound Med 1989;8:149-153. 12. Hennerici M, Steinke W, Rautenberg W. High-resistance Doppler flow pattern in extracranial carotid dissection. Arch Neurol 1989; 46:670-672. 13. Rothrock JF, Lim V, Press G, Gosink B. Serial magnetic resonance and carotid duplex examinations in the management of carotid dissection. Neurology 1989;39:686-692. 14. O’Leary DH, Polak JF, Kronmal RA, et al. Carotid artery intima and media thickness as a risk factor for myocardial infarction and stroke in older adults. Cardiovascular Health Study Collaborative Research Group. N Engl J Med 1999;340:14-22. 15. Sidhu PS, Jonker ND, Khaw KT, et al. Spontaneous dissections of the internal carotid artery: appearances on colour Doppler ultrasound. Br J Radiol 1997;70:50-57. 16. Gritzmann N, Grasl MC, Helmer M, Steiner E. Invasion of the carotid artery and jugular vein by lymph node metastases: detection with sonography. AJR Am J Roentgenol 1990;154:411-414. 17. Gooding GA, Langman AW, Dillon WP, Kaplan MJ. Malignant carotid artery invasion: sonographic detection. Radiology 1989;171: 435-438. 18. Steinke W, Hennerici M, Aulich A. Doppler color flow imaging of carotid body tumors. Stroke 1989;20:1574-1577. 19. Grant EG, Wong W, Tessler F, Perrella R. Cerebrovascular ultrasound imaging. Radiol Clin North Am 1988;26:1111-1130. Carotid Ultrasound Examination 20. Ide C, De Coene B, Mailleux P, et al. Hypoplasia of the internal carotid artery: a noninvasive diagnosis. Eur Radiol 2000;10:18651870. Carotid Ultrasound Interpretation 21. Polak JF, O’Leary DH, Kronmal RA, et al. Sonographic evaluation of carotid artery atherosclerosis in the elderly: relationship of disease severity to stroke and transient ischemic attack. Radiology 1993;188: 363-370. 22. Veller MG, Fisher CM, Nicolaides AN, et al. Measurement of the ultrasonic intima-media complex thickness in normal subjects. J Vasc Surg 1993;17:719-725. 23. Bots ML, Mulder PG, Hofman A, et al. Reproducibility of carotid vessel wall thickness measurements: the Rotterdam Study. J Clin Epidemiol 1994;47:921-930. 24. Csanyi A, Egervari A. Simple clinical method of average intimamedia thickness measurement in the common carotid artery. Vasa 1996;25:242-248. 25. Bots ML, Hoes AW, Koudstaal PJ, et al. Common carotid intimamedia thickness and risk of stroke and myocardial infarction: the Rotterdam Study. Circulation 1997;96:1432-1437. 26. Kanters SD, Algra A, van Leeuwen MS, Banga JD. Reproducibility of in vivo carotid intima-media thickness measurements: a review. Stroke 1997;28:665-671. 27. Dwyer JH, Sun P, Kwong-Fu H, et al. Automated intima-media thickness: the Los Angeles Atherosclerosis Study. Ultrasound Med Biol 1998;24:981-987. 28. Aminbakhsh A, Frohlich J, Mancini GB. Detection of early atherosclerosis with B mode carotid ultrasonography: assessment of a new quantitative approach. Clin Invest Med 1999;22:265-274. 29. Greenland P, Abrams J, Aurigemma GP, et al. Prevention Conference V. Beyond secondary prevention: identifying the high-risk patient for primary prevention—noninvasive tests of atherosclerotic burden. Writing Group III. Circulation 2000;101:E16-E22. 30. Bluth EI, Stavros AT, Marich KW, et al. Carotid duplex sonography: a multicenter recommendation for standardized imaging and Doppler criteria. Radiographics 1988;8:487-506. 31. Bluth EI. Evaluation and characterization of carotid plaque. Semin Ultrasound CT MR 1997;18:57-65. 32. Polak JF, Shemanski L, O’Leary DH, et al. Hypoechoic plaque at

ultrasound of the carotid artery: an independent risk factor for incident stroke in adults aged 65 years or older. Cardiovascular Health Study. Radiology 1998;208:649-654. 33. Langsfeld M, Gray-Weale AC, Lusby RJ. The role of plaque morphology and diameter reduction in the development of new symptoms in asymptomatic carotid arteries. J Vasc Surg 1989;9:548557. 34. Leahy AL, McCollum PT, Feeley TM, et al. Duplex ultrasonography and selection of patients for carotid endarterectomy: plaque morphology or luminal narrowing? J Vasc Surg 1988;8:558-562. 35. Reilly LM, Lusby RJ, Hughes L, et al. Carotid plaque histology using real-time ultrasonography: clinical and therapeutic implications. Am J Surg 1983;146:188-193. 36. Persson AV, Robichaux WT, Silverman M. The natural history of carotid plaque development. Arch Surg 1983;118:1048-1052. 37. Lusby RJ, Ferrell LD, Ehrenfeld WK, et al. Carotid plaque hemorrhage: its role in production of cerebral ischemia. Arch Surg 1982;117:1479-1488. 38. Edwards JH, Kricheff II, Gorstein F, et al. Atherosclerotic sub­ intimal hematoma of the carotid artery. Radiology 1979;133: 123-129. 39. Imparato AM, Riles TS, Gorstein F. The carotid bifurcation plaque: pathologic findings associated with cerebral ischemia. Stroke 1979; 10:238-245. 40. Geroulakos G, Ramaswami G, Nicolaides A, et al. Characterization of symptomatic and asymptomatic carotid plaques using high-resolution real-time ultrasonography. Br J Surg 1993;80:1274-1277. 41. Holdsworth RJ, McCollum PT, Bryce JS, Harrison DK. Symptoms, stenosis and carotid plaque morphology: is plaque morphology relevant? Eur J Vasc Endovasc Surg 1995;9:80-85. 42. Bluth EI, Kay D, Merritt CR, et al. Sonographic characterization of carotid plaque: detection of hemorrhage. AJR Am J Roentgenol 1986;146:1061-1065. 43. Merritt CRB, Bluth EI. Ultrasonographic characterization of carotid plaque. In: Labs KH, editor. Diagnostic vascular ultrasound. London: Hodder & Stoughton; 1991. 44. Bluth EI. B-mode evaluation and characterization of carotid plaque. In: Tegler CH, Babikian VL, Gomez CR, editors. Neurosonology. St Louis: Mosby–Year Book; 1991. p. 62-67. 45. Sterpetti AV, Schultz RD, Feldhaus RJ, et al. Ultrasonographic features of carotid plaque and the risk of subsequent neurologic deficits. Surgery 1988;104:652-660. 46. Weinberger J, Marks SJ, Gaul JJ, et al. Atherosclerotic plaque at the carotid artery bifurcation: correlation of ultrasonographic imaging with morphology. J Ultrasound Med 1987;6:363-366. 47. Stahl JA, Middleton WD. Pseudoulceration of the carotid artery. J Ultrasound Med 1992;11:355-358. 48. Bluth EI. Extracranial carotid arteries: intraplaque hemorrhage and surface ulceration. Minerva Cardioangiol 1998;46:81-85. 49. Bluth EI. Plaque morphology as a risk factor for stroke. JAMA 2000;284:177. 50. Reiter M, Effenberger I, Sabeti S, et al. Increasing carotid plaque echolucency is predictive of cardiovascular events in high-risk patients. Radiology 2008;248:1050-1055. 51. Bluth EI, McVay 3rd LV, Merritt CR, Sullivan MA. The identification of ulcerative plaque with high-resolution duplex carotid scanning. J Ultrasound Med 1988;7:73-76. 52. Ballard JL, Deiparine MK, Bergan JJ, et al. Cost-effective evaluation and treatment for carotid disease. Arch Surg 1997;132:268271. 53. Furst H, Hartl WH, Jansen I, et al. Color-flow Doppler sonography in the identification of ulcerative plaques in patients with high-grade carotid artery stenosis. AJNR Am J Neuroradiol 1992;13: 1581-1587. 54. Abildgaard A, Egge TS, Klow NE, Jakobsen JA. Use of sonicated albumin (Infoson) to enhance arterial spectral and color Doppler imaging. Cardiovasc Intervent Radiol 1996;19:265-271. 55. Bluth EI. Value of ultrasound in selecting patients for carotid angioplasty and stent placement. Radiology 2005;237:374-375; author reply 375. 56. Comerota AJ, Cranley JJ, Cook SE. Real-time B-mode carotid imaging in diagnosis of cerebrovascular disease. Surgery 1981;89: 718-729. 57. Zwiebel WJ, Austin CW, Sackett JF, Strother CM. Correlation of high-resolution, B-mode and continuous-wave Doppler sonography with arteriography in the diagnosis of carotid stenosis. Radiology

Chapter 25  ■  The Extracranial Cerebral Vessels   995 1983;149:523-532. 58. Jacobs NM, Grant EG, Schellinger D, et al. Duplex carotid sonography: criteria for stenosis, accuracy, and pitfalls. Radiology 1985; 154:385-391. 59. Taylor KJ, Holland S. Doppler ultrasound. Part I. Basic principles, instrumentation, and pitfalls. Radiology 1990;174:297-307. 60. Carroll BA, von Ramm OT. Fundamental of current Doppler technology. Ultrasound Q 1988;6:275-298. 61. Kassam M, Johnston KW, Cobbold RS. Quantitative estimation of spectral broadening for the diagnosis of carotid arterial disease: method and in vitro results. Ultrasound Med Biol 1985;11: 425-433. 62. Douville Y, Johnston KW, Kassam M. Determination of the hemodynamic factors which influence the carotid Doppler spectral broadening. Ultrasound Med Biol 1985;11:417-423. 63. Garth KE, Carroll BA, Sommer FG, Oppenheimer DA. Duplex ultrasound scanning of the carotid arteries with velocity spectrum analysis. Radiology 1983;147:823-827. 64. Phillips DJ, Greene Jr FM, Langlois Y, et al. Flow velocity patterns in the carotid bifurcations of young, presumed normal subjects. Ultrasound Med Biol 1983;9:39-49. 65. Lichtman JB, Kibble MB. Detection of intracranial arteriovenous malformation by Doppler ultrasound of the extracranial carotid circulation. J Ultrasound Med 1987;6:609-612. 66. Robinson ML, Sacks D, Perlmutter GS, Marinelli DL. Diagnostic criteria for carotid duplex sonography. AJR Am J Roentgenol 1988;151:1045-1049. 67. Grant EG, Duerinckx AJ, El Saden SM, et al. Ability to use duplex ultrasound to quantify internal carotid arterial stenoses: fact or fiction? Radiology 2000;214:247-252. 68. Rothwell PM, Gibson RJ, Slattery J, Warlow CP. Prognostic value and reproducibility of measurements of carotid stenosis: a comparison of three methods on 1001 angiograms. European Carotid Surgery Trialists’ Collaborative Group. Stroke 1994;25:24402444. 69. Moneta GL, Edwards JM, Chitwood RW, et al. Correlation of North American Symptomatic Carotid Endarterectomy Trial (NASCET) angiographic definition of 70% to 99% internal carotid artery stenosis with duplex scanning. J Vasc Surg 1993;17:152-157; discussion 157-159. 70. Friedman SG, Hainline B, Feinberg AW, et al. Use of diastolic velocity ratios to predict significant carotid artery stenosis. Stroke 1988;19:910-912. 71. Kohler TR, Langlois Y, Roederer GO, et al. Variability in measurement of specific parameters for carotid duplex examination. Ultrasound Med Biol 1987;13:637-642. 72. Hunink MG, Polak JF, Barlan MM, O’Leary DH. Detection and quantification of carotid artery stenosis: efficacy of various Doppler velocity parameters. AJR Am J Roentgenol 1993;160:619-625. 73. Horrow MM, Stassi J, Shurman A, et al. The limitations of carotid sonography: interpretive and technology-related errors. AJR Am J Roentgenol 2000;174:189-194. 74. Faught WE, Mattos MA, van Bemmelen PS, et al. Color-flow duplex scanning of carotid arteries: new velocity criteria based on receiver operator characteristic analysis for threshold stenoses used in the symptomatic and asymptomatic carotid trials. J Vasc Surg 1994;19:818-827; discussion 827-828. 75. Kuntz KM, Polak JF, Whittemore AD, et al. Duplex ultrasound criteria for the identification of carotid stenosis should be laboratory specific. Stroke 1997;28:597-602. 76. Alexandrov AV, Vital D, Brodie DS, et al. Grading carotid stenosis with ultrasound: an interlaboratory comparison. Stroke 1997;28: 1208-1210. 77. Society of Radiologists in Ultrasound. Consensus Conference on Carotid Ultrasound. October 2002. Radiology 2003;229:340-346. 78. Carpenter JP, Lexa FJ, Davis JT. Determination of sixty percent or greater carotid artery stenosis by duplex Doppler ultrasonography. J Vasc Surg 1995;22:697-703; discussion 703-705. 79. Middleton WD, Erickson S, Melson GL. Perivascular color artifact: pathologic significance and appearance on color Doppler ultrasound images. Radiology 1989;171:647-652. 80. Erickson SJ, Mewissen MW, Foley WD, et al. Stenosis of the internal carotid artery: assessment using color Doppler imaging compared with angiography. AJR Am J Roentgenol 1989;152: 1299-1305. 81. Middleton WD, Foley WD, Lawson TL. Flow reversal in the normal

carotid bifurcation: color Doppler flow imaging analysis. Radiology 1988;167:207-210. 82. Zierler RE, Phillips DJ, Beach KW, et al. Noninvasive assessment of normal carotid bifurcation hemodynamics with color-flow ultrasound imaging. Ultrasound Med Biol 1987;13:471-476. 83. Erickson SJ, Mewissen MW, Foley WD, et al. Color Doppler evaluation of arterial stenoses and occlusions involving the neck and thoracic inlet. Radiographics 1989;9:389-406. 84. Middleton WD, Foley WD, Lawson TL. Color-flow Doppler imaging of carotid artery abnormalities. AJR Am J Roentgenol 1988;150:419-425. 85. Branas CC, Weingarten MS, Czeredarczuk M, Schafer PF. Examination of carotid arteries with quantitative color Doppler flow imaging. J Ultrasound Med 1994;13:121-127. 86. Steinke W, Ries S, Artemis N, et al. Power Doppler imaging of carotid artery stenosis: comparison with color Doppler flow imaging and angiography. Stroke 1997;28:1981-1987. 87. Bluth EI, Althans LE, Sullivan M, et al. Comparison of plaque characterization with grayscale imaging and 3-D power Doppler imaging: can more be learned about intraplaque hemorrhage? JEMU 1999;20:11-15. 88. Griewing B, Morgenstern C, Driesner F, et al. Cerebrovascular disease assessed by color-flow and power Doppler ultrasonography: comparison with digital subtraction angiography in internal carotid artery stenosis. Stroke 1996;27:95-100. 89. Bluth EI, Sunshine JH, Lyons JB, et al. Power Doppler imaging: initial evaluation as a screening examination for carotid artery stenosis. Radiology 2000;215:791-800. 90. Bluth EI. Screening test for carotid disease. Semin Ultrasound CT MR 2003;24:55-61. 91. Zbornikova V, Lassvik C. Duplex scanning in presumably normal persons of different ages. Ultrasound Med Biol 1986;12:371-378. 92. Spencer EB, Sheafor DH, Hertzberg BS, et al. Nonstenotic internal carotid arteries: effects of age and blood pressure at the time of scanning on Doppler ultrasound velocity measurements. Radiology 2001;220:174-178. 93. O’Boyle MK, Vibhakar NI, Chung J, et al. Duplex sonography of the carotid arteries in patients with isolated aortic stenosis: imaging findings and relation to severity of stenosis. AJR Am J Roentgenol 1996;166:197-202. 94. Macchi C, Gulisano M, Giannelli F, et al. Kinking of the human internal carotid artery: a statistical study in 100 healthy subjects by echocolor Doppler. J Cardiovasc Surg (Torino) 1997;38:629-637. 95. Busuttil SJ, Franklin DP, Youkey JR, Elmore JR. Carotid duplex overestimation of stenosis due to severe contralateral disease. Am J Surg 1996;172:144-147; discussion 147-148. 96. AbuRahma AF, Richmond BK, Robinson PA, et al. Effect of contralateral severe stenosis or carotid occlusion on duplex criteria of ipsilateral stenoses: comparative study of various duplex parameters. J Vasc Surg 1995;22:751-761; discussion 761-762. 97. Van Everdingen KJ, van der Grond J, Kappelle LJ. Overestimation of a stenosis in the internal carotid artery by duplex sonography caused by an increase in volume flow. J Vasc Surg 1998;27: 479-485. 98. Blackshear WM, Phillips DJ, Chikos PM, et al. Carotid artery velocity patterns in normal and stenotic vessels. Stroke 1980; 11:67-71. 99. Grubb Jr RL, Derdeyn CP, Fritsch SM, et al. Importance of hemodynamic factors in the prognosis of symptomatic carotid occlusion. JAMA 1998;280:1055-1060. 100. Berman SS, Devine JJ, Erdoes LS, Hunter GC. Distinguishing carotid artery pseudo-occlusion with color-flow Doppler. Stroke 1995;26:434-438. 101. Gortler M, Niethammer R, Widder B. Differentiating subtotal carotid artery stenoses from occlusions by colour-coded duplex sonography. J Neurol 1994;241:301-305. 102. AbuRahma AF, Pollack JA, Robinson PA, Mullins D. The reliability of color duplex ultrasound in diagnosing total carotid artery occlusion. Am J Surg 1997;174:185-187. 103. Kliewer MA, Freed KS, Hertzberg BS, et al. Temporal artery tap: usefulness and limitations in carotid sonography. Radiology 1996; 201:481-484. 104. Bebry AJ, Hines GL. Total occlusion of the common carotid artery with a patent internal carotid artery: identification by duplex ultrasonography—report of a case. J Vasc Surg 1989;10:469-470. 105. Blackshear Jr WM, Phillips DJ, Bodily KC, Strandness Jr DE.

996   PART III  ■  Small Parts, Carotid Artery, and Peripheral Vessel Sonography Ultrasonic demonstration of external and internal carotid patency with common carotid occlusion: a preliminary report. Stroke 1980; 11:249-252. 106. Lee DH, Gao FQ, Rankin RN, et al. Duplex and color Doppler flow sonography of occlusion and near occlusion of the carotid artery. AJNR Am J Neuroradiol 1996;17:1267-1274. 107. Alexandrov AV, Bladin CF, Maggisano R, Norris JW. Measuring carotid stenosis: time for a reappraisal. Stroke 1993;24:1292-1296. 108. Polak JF, Kalina P, Donaldson MC, et al. Carotid endarterectomy: preoperative evaluation of candidates with combined Doppler sonography and MR angiography—work in progress. Radiology 1993;186:333-338. 109. Johnston DC, Goldstein LB. Clinical carotid endarterectomy decision making: noninvasive vascular imaging versus angiography. Neurology 2001;56:1009-1015. 110. Kuntz KM, Skillman JJ, Whittemore AD, Kent KC. Carotid endarterectomy in asymptomatic patients–is contrast angiography necessary? A morbidity analysis. J Vasc Surg 1995;22:706-714; discussion 714-716. 111. Mattos MA, Hodgson KJ, Faught WE, et al. Carotid endarterectomy without angiography: is color-flow duplex scanning sufficient? Surgery 1994;116:776-782; discussion 782-783. 112. Cartier R, Cartier P, Fontaine A. Carotid endarterectomy without angiography: the reliability of Doppler ultrasonography and duplex scanning in preoperative assessment. Can J Surg 1993;36:411-416. 113. Thusay MM, Khoury M, Greene K. Carotid endarterectomy based on duplex ultrasound in patients with and without hemispheric symptoms. Am Surg 2001;67:1-6. 114. Welch HJ, Murphy MC, Raftery KB, Jewell ER. Carotid duplex with contralateral disease: the influence of vertebral artery blood flow. Ann Vasc Surg 2000;14:82-88. 115. Chen JC, Salvian AJ, Taylor DC, et al. Can duplex ultrasonography select appropriate patients for carotid endarterectomy? Eur J Vasc Endovasc Surg 1997;14:451-456. 116. Randoux B, Marro B, Koskas F, et al. Carotid artery stenosis: prospective comparison of CT, three-dimensional gadolinium-enhanced MR, and conventional angiography. Radiology 2001;220:179-185. 117. Johnson BL, Gupta AK, Bandyk DF, et al. Anatomic patterns of carotid endarterectomy healing. Am J Surg 1996;172:188-190. 118. Kagawa R, Okada Y, Shima T, et al. B-mode ultrasonographic investigations of morphological changes in endarterectomized carotid artery. Surg Neurol 2001;55:50-56; discussion 56-57. 119. Jackson MR, D’Addio VJ, Gillespie DL, O’Donnell SD. The fate of residual defects following carotid endarterectomy detected by early postoperative duplex ultrasound. Am J Surg 1996;172: 184-187. 120. Ricotta JJ, DeWeese JA. Is routine carotid ultrasound surveillance after carotid endarterectomy worthwhile? Am J Surg 1996;172:140142; discussion 143. 121. Goodney PP, Lucas FL, Likosky DS, et al. Changes in the use of carotid revascularization among the Medicare population. Arch Surg 2008;143:170-173. 122. Chaer RA, Derubertis BG, Trocciola SM, et al. Safety and efficacy of carotid angioplasty and stenting in high-risk patients. Am Surg 2006;72:694-698; discussion 698-699. 123. Wolff T, Guirguis-Blake J, Miller T, et al. Screening for carotid artery stenosis: an update of the evidence for the U.S. Preventive Services Task Force. Ann Intern Med 2007;147:860-870. 124. Diethrich EB, Pauliina Margolis M, Reid DB, et al. Virtual histology intravascular ultrasound assessment of carotid artery disease: the Carotid Artery Plaque Virtual Histology Evaluation (CAPITAL) study. J Endovasc Ther 2007;14:676-686. 125. Reiter M, Bucek RA, Effenberger I, et al. Plaque echolucency is not associated with the risk of stroke in carotid stenting. Stroke 2006;37:2378-2380. 126. Fleming SE, Bluth EI, Milburn J. Role of sonography in the evaluation of carotid artery stents. J Clin Ultrasound 2005; 33:321-328. 127. Peterson BG, Longo GM, Kibbe MR, et al. Duplex ultrasound remains a reliable test even after carotid stenting. Ann Vasc Surg 2005;19:793-797. 128. Roffi M, Greutmann M, Eberli FR, et al. Starting a carotid artery stenting program is safe. Catheter Cardiovasc Interv 2008;71: 469-473. 129. Zhou W, Lin PH, Bush RL, et al. Management of in-sent restenosis

after carotid artery stenting in high-risk patients. J Vasc Surg 2006;43:305-312. 130. Robbin ML, Lockhart ME, Weber TM, et al. Carotid artery stents: early and intermediate follow-up with Doppler ultrasound. Radiology 1997;205:749-756. 131. Chahwan S, Miller MT, Pigott JP, et al. Carotid artery velocity characteristics after carotid artery angioplasty and stenting. J Vasc Surg 2007;45:523-526. 132. Setacci C, Chisci E, Setacci F, et al. Grading carotid intrastent restenosis: a 6-year follow-up study. Stroke 2008;39:11891196. 133. Zhou W, Felkai DD, Evans M, et al. Ultrasound criteria for severe in-stent restenosis following carotid artery stenting. J Vasc Surg 2008;47:74-80. 134. Lal BK, Hobson 2nd RW, Tofighi B, et al. Duplex ultrasound velocity criteria for the stented carotid artery. J Vasc Surg 2008; 47:63-73. 135. Armstrong PA, Bandyk DF, Johnson BL, et al. Duplex scan surveillance after carotid angioplasty and stenting: a rational definition of stent stenosis. J Vasc Surg 2007;46:460-465; discussion 465-466. 136. Chi YW, White CJ, Woods TC, Goldman CK. Ultrasound velocity criteria for carotid in-stent restenosis. Catheter Cardiovasc Interv 2007;69:349-354. Nonatherosclerotic Carotid Disease 137. Furie DM, Tien RD. Fibromuscular dysplasia of arteries of the head and neck: imaging findings. AJR Am J Roentgenol 1994;162: 1205-1209. 138. Kliewer MA, Carroll BA. Ultrasound case of the day: internal carotid artery web (atypical fibromuscular dysplasia). Radiographics 1991; 11:504-505. 139. Maeda H, Handa N, Matsumoto M, et al. Carotid lesions detected by B-mode ultrasonography in Takayasu’s arteritis: “macaroni sign” as an indicator of the disease. Ultrasound Med Biol 1991;17: 695-701. 140. Sturzenegger M. Spontaneous internal carotid artery dissection: early diagnosis and management in 44 patients. J Neurol 1995;242: 231-238. 141. Sturzenegger M, Mattle HP, Rivoir A, Baumgartner RW. Ultrasound findings in carotid artery dissection: analysis of 43 patients. Neurology 1995;45:691-698. 142. Steinke W, Rautenberg W, Schwartz A, Hennerici M. Noninvasive monitoring of internal carotid artery dissection. Stroke 1994;25: 998-1005. Transcranial Doppler Sonography 143. Lupetin AR, Davis DA, Beckman I, Dash N. Transcranial Doppler sonography. Part 1. Principles, technique, and normal appearances. Radiographics 1995;15:179-191. 144. Comerota AJ, Katz ML, Hosking JD, et al. Is transcranial Doppler a worthwhile addition to screening tests for cerebrovascular disease? J Vasc Surg 1995;21:90-95; discussion 95-97. 145. Rorick MB, Nichols FT, Adams RJ. Transcranial Doppler correlation with angiography in detection of intracranial stenosis. Stroke 1994;25:1931-1934. 146. Ultrasound screening helps prevent stroke in children with sickle cell disease. Science Centric 7 Dec 2008. 147. Lupetin AR, Davis DA, Beckman I, Dash N. Transcranial Doppler sonography. Part 2. Evaluation of intracranial and extracranial abnormalities and procedural monitoring. Radiographics 1995;15: 193-209. 148. Lin SK, Ryu SJ, Chu NS. Carotid duplex and transcranial colorcoded sonography in evaluation of carotid-cavernous sinus fistulas. J Ultrasound Med 1994;13:557-564. 149. Mast H, Mohr JP, Thompson JL, et al. Transcranial Doppler ultrasonography in cerebral arteriovenous malformations: diagnostic sensitivity and association of flow velocity with spontaneous hemorrhage and focal neurological deficit. Stroke 1995;26: 1024-1027. 150. Gaunt ME, Martin PJ, Smith JL, et al. Clinical relevance of intraoperative embolization detected by transcranial Doppler ultrasonography during carotid endarterectomy: a prospective study of 100 patients. Br J Surg 1994;81:1435-1439.

Chapter 25  ■  The Extracranial Cerebral Vessels   997 Vertebral Artery 151. Bendick PJ, Glover JL. Hemodynamic evaluation of vertebral arteries by duplex ultrasound. Surg Clin North Am 1990;70:235-244. 152. Lewis BD, James EM, Welch TJ. Current applications of duplex and color Doppler ultrasound imaging: carotid and peripheral vascular system. Mayo Clin Proc 1989;64:1147-1157. 153. Visona A, Lusiani L, Castellani V, et al. The echo-Doppler (duplex) system for the detection of vertebral artery occlusive disease: comparison with angiography. J Ultrasound Med 1986;5:247-250. 154. Davis PC, Nilsen B, Braun IF, Hoffman Jr JC. A prospective comparison of duplex sonography vs angiography of the vertebral arteries. AJNR Am J Neuroradiol 1986;7:1059-1064. 155. Bluth EI, Merritt CR, Sullivan MA, et al. Usefulness of duplex ultrasound in evaluating vertebral arteries. J Ultrasound Med 1989; 8:229-235. 156. Walker DW, Acker JD, Cole CA. Subclavian steal syndrome detected with duplex pulsed Doppler sonography. AJNR Am J Neuroradiol 1982;3:615-618. 157. Elias DA, Weinberg PE. Angiography of the posterior fossa. In: Taveras JM, Ferrucci JT, editors. Radiology: diagnosis-imagingintervention. Philadelphia: Lippincott; 1989. 158. Bendick PJ, Jackson VP. Evaluation of the vertebral arteries with duplex sonography. J Vasc Surg 1986;3:523-530. 159. Ackerstaff RG, Grosveld WJ, Eikelboom BC, Ludwig JW. Ultrasonic duplex scanning of the prevertebral segment of the vertebral artery in patients with cerebral atherosclerosis. Eur J Vasc Surg 1988;2:387-393. 160. Carroll BA, Holder CA. Vertebral artery duplex sonography (abstract). J Ultrasound Med 1990;9:S27-S28. 161. De Bray JM, Zenglein JP, Laroche JP, et al. Effect of subclavian syndrome on the basilar artery. Acta Neurol Scand 1994;90: 174-178. 162. Thomassen L, Aarli JA. Subclavian steal phenomenon: clinical and hemodynamic aspects. Acta Neurol Scand 1994;90:241-244. 163. Kliewer MA, Hertzberg BS, Kim DH, et al. Vertebral artery Doppler waveform changes indicating subclavian steal physiology. AJR Am J Roentgenol 2000;174:815-819. 164. Nicolau C, Gilabert R, Garcia A, et al. Effect of internal carotid artery occlusion on vertebral artery blood flow: a duplex ultrasonographic evaluation. J Ultrasound Med 2001;20:105-111. Internal Jugular Veins 165. Williams CE, Lamb GH, Roberts D, Davies J. Venous thrombosis in the neck: the role of real-time ultrasound. Eur J Radiol 1989;9: 32-36.

166. Hubsch PJ, Stiglbauer RL, Schwaighofer BW, et al. Internal jugular and subclavian vein thrombosis caused by central venous catheters: evaluation using Doppler blood flow imaging. J Ultrasound Med 1988;7:629-636. 167. Gaitini D, Kaftori JK, Pery M, Engel A. High-resolution real-time ultrasonography: diagnosis and follow-up of jugular and subclavian vein thrombosis. J Ultrasound Med 1988;7:621-627. 168. Albertyn LE, Alcock MK. Diagnosis of internal jugular vein thrombosis. Radiology 1987;162:505-508. 169. Falk RL, Smith DF. Thrombosis of upper extremity thoracic inlet veins: diagnosis with duplex Doppler sonography. AJR Am J Roentgenol 1987;149:677-682. 170. Weissleder R, Elizondo G, Stark DD. Sonographic diagnosis of subclavian and internal jugular vein thrombosis. J Ultrasound Med 1987;6:577-587. 171. De Witte BR, Lameris JS. Real-time ultrasound diagnosis of internal jugular vein thrombosis. J Clin Ultrasound 1986;14:712-717. 172. Wing V, Scheible W. Sonography of jugular vein thrombosis. AJR Am J Roentgenol 1983;140:333-336. 173. Gribbin C, Raghavendra BN, Ginsburg HB. Ultrasound diagnosis of jugular venous ectasia. NY State J Med 1989;89:532-533. 174. Hughes PL, Qureshi SA, Galloway RW. Jugular venous aneurysm in children. Br J Radiol 1988;61:1082-1084. 175. Jasinski RW, Rubin JM. CT and ultrasonographic findings in jugular vein ectasia. J Ultrasound Med 1984;3:417-420. 176. Stevens RK, Fried AM, Hood Jr TR. Ultrasonic diagnosis of jugular venous aneurysm. J Clin Ultrasound 1982;10:85-87. 177. Lee W, Leduc L, Cotton DB. Ultrasonographic guidance for central venous access during pregnancy. Am J Obstet Gynecol 1989;161: 1012-1023. 178. Bond DM, Champion LK, Nolan R. Real-time ultrasound imaging aids jugular venipuncture. Anesth Analg 1989;68:700-701. 179. Machi J, Takeda J, Kakegawa T. Safe jugular and subclavian venipuncture under ultrasonographic guidance. Am J Surg 1987;153: 321-323. 180. Dresser LP, McKinney WM. Anatomic and pathophysiologic studies of the human internal jugular valve. Am J Surg 1987;154: 220-224. 181. Patel S, Brennan J. Diagnosis of internal jugular vein thrombosis by computed tomography. J Comput Assist Tomogr 1981;5:197-200. 182. Braun IF, Hoffman Jr JC, Malko JA, et al. Jugular venous thrombosis: MR imaging. Radiology 1985;157:357-360.

CHAPTER 26 

The Peripheral Arteries Joseph F. Polak and Jean M. Alessi-Chinetti

Chapter Outline DIAGNOSTIC SCREENING METHODS SONOGRAPHIC TECHNIQUE Real-Time Gray-Scale Imaging Doppler Sonography DOPPLER FLOW PATTERNS Normal Arteries Stenotic Arteries Arteriovenous Fistulas Masses PERIPHERAL ARTERY DISEASE Incidence and Clinical Importance Sonographic Technique

Lower Extremity

Normal Anatomy Aneurysms Stenoses and Occlusions

Upper Extremity

Normal Anatomy and Doppler Flow Patterns Pathophysiology and Diagnostic Accuracy

VASCULAR AND PERIVASCULAR MASSES Synthetic Vascular Bypass Grafts Masses: Hematoma versus Pseudoaneurysm

DIAGNOSTIC SCREENING METHODS The upper and lower extremity arteries are readily evaluated by Doppler ultrasound. Because they are usually located at depths of 6 cm or less, the extremity arteries are more consistently imaged than the abdominal or thoracic arteries. Availability of sufficient imaging windows allows the transducer to be placed over the artery of interest without the presence of overlying attenuating tissues containing bone or gas. Transducers with imaging frequencies greater than 5 MHz can normally be used because the arteries lie close to the skin. Doppler frequencies are typically more than 3 MHz. Although limited, real-time gray-scale sonography is useful for evaluating the presence of atherosclerotic plaque or confirming the presence of extravascular masses. Color flow Doppler sonographic imaging allows the clinician to survey the area of interest rapidly, determine if vascular structures are present, and if so, characterize their blood flow patterns (Fig. 26-1; Video 26-1). The addition of spectral Doppler waveform analysis makes duplex Doppler sonography a powerful diagnostic tool for evaluating the clinical significance of atherosclerotic lesions, differentiating significant arterial stenoses from occlusions and assessing the nature of perivascular masses. Compared with duplex sonography (spectral Doppler and gray-scale sonography) alone, color flow Doppler imaging can more rapidly survey the full length 998

Occlusions and Anastomotic Stenoses AUTOLOGOUS VEIN GRAFTS Stenosis of Venous Bypass Graft Arteriovenous Fistula DIALYSIS ACCESS GRAFTS AND FISTULAS COMPLICATIONS OF INVASIVE PROCEDURES Fistulous Communications Pseudoaneurysms CLOSURE DEVICES CONCLUSION

of limb arteries and detect the presence of significant stenoses and occlusions. Color flow imaging decreases the length of the peripheral arterial examination compared with duplex sonography alone,1 and it improves diagnostic accuracy.2 As such, peripheral artery imaging is, in fact, color Doppler sonography. Power Doppler sonography, a more sensitive derivative of color Doppler imaging, can further improve the diagnostic performance of Doppler sonography in specific clinical situations. Compared with angiography, the sonographic approaches discussed in this chapter have the advantage of being noninvasive, relatively inexpensive, and well suited for serial examinations. They also permit the evaluation of soft tissue structures contiguous to the arteries. Computed tomography angiography (CTA) is a more expensive technology than Doppler sonography and requires the administration of contrast material. The use of multidetector CTA has shortened imaging times, improved resolution, and made it competitive with arteriography. Magnetic resonance angiography (MRA) is also used to detect the presence of arterial lesions. As with CTA, MRA can also be used to evaluate the soft tissues for nonvascular pathologies. However, MRA requires additional imaging sequences and increases imaging time. CTA and MRA are less operator dependent and, in given clinical situations, more accurate and more reproducible than Doppler sonography. However, the value of Doppler ultrasound is undisputed in patients with poor renal func-

Chapter 26  ■  The Peripheral Arteries   999

RT SFA PROX TO MID THIGH

A

B

FIGURE 26-1.  Stenosis of superficial femoral artery on color and spectral Doppler ultrasound. A, Color blood flow image shows that the site of stenosis causes an alteration in the color signals within the artery. There is aliasing of the color Doppler signals (blue) at the site of maximal stenosis. Abnormalities in the color flow signals extend at least 1 cm downstream from the lesion. Calcified plaques cause areas of Doppler signal drop-off distal to the stenosis. B, Pulsed Doppler waveforms taken before and at the site of aliasing show a significant increase in velocity (doubling), consistent with a significant stenosis of the superficial femoral artery.

tion. Although MRA and CTA are more cost-effective than sonography in certain clinical scenarios, Doppler ultrasound can often help identify patients requiring direct referral for percutaneous arterial interventions.3-6

SONOGRAPHIC TECHNIQUE Real-Time Gray-Scale Imaging The diameter of the peripheral arteries that is clinically relevant varies from 1 to 6 mm. Accurate visualization of the arterial wall requires high-resolution transducers, more than 5 MHz, to visualize all the various lesions. A broad frequency range of 5 to 10 MHz is preferred because it offers overall good resolution while permitting good depth penetration, even in the thigh. For detailed visualization of smaller-diameter arteries, higher frequencies of 7 to 12 MHz can be used. At these high frequencies, transducers have poor depth penetration but may be useful for evaluating bypass grafts and the ulnar and radial arteries and smaller arteries of the hand. The linear phased array transducer is ideal for imaging the extremity arteries. The transducer has sufficient length to permit rapid coverage of long arterial segments by holding it parallel to the artery or graft long axis and by sliding it in a series of nonoverlapping increments. A smaller-footprint, curved array or sector transducer can be useful for imaging the iliac arteries and the more centrally located portions of the subclavian arteries.

Doppler Sonography Simultaneous display of Doppler spectral waveforms and data and the gray-scale image, duplex Doppler

sonography, is the basic requisite for the evaluation of the peripheral arteries and arterial bypass grafts.7 Careful real-time control is needed to position the Doppler sample gate and accurately detect sites of maximal blood flow velocity in arteries and bypass grafts. The transducer carrier frequencies can vary between 3 to 10 MHz, tending to be best lower than the simultaneously acquired gray-scale image. Selection of a Doppler transducer frequency of approximately 5 MHz sacrifices some sensi­ tivity for detecting slowly moving blood, but decreases the likelihood that the system will alias at sites of rapidly moving blood, such as stenoses or arteriovenous fistulas. Color Doppler sonography is an essential component of a peripheral arterial sonographic examination. The simultaneous display of moving blood superimposed on a gray-scale image allows a rapid survey of the flow patterns within long sections of the peripheral arteries and bypass grafts.8 In general, an efficient approach to peripheral vascular sonography relies on color flow Doppler sonography to rapidly identify zones of flow disturbances, then on duplex sonography, including Doppler spectral analysis, to characterize the type of flow abnormality present.1,9 The color Doppler image displays only the mean frequency shift caused by moving structures. The pixel size (resolution) is also coarser than the corresponding pixel size of gray-scale image. This may cause some ambiguity in alignment of the two separate images and can cause the color Doppler information to overlap beyond the wall of the arteries. Most manufacturers use lower transducer frequencies for the color flow image than for the gray-scale sonographic component of the image. This approach increases the depth penetration of the color flow image without compromising image resolution.

1000   PART III  ■  Small Parts, Carotid Artery, and Peripheral Vessel Sonography

Power Doppler sonography is a variant of color flow Doppler imaging that displays a summation of the Doppler signals caused by moving blood. Advantages of power Doppler over color Doppler flow imaging are (1) the blood flow information does not alias, (2) the signal strengths are much less angle dependent, and (3) slowly moving blood is more easily detected. A disadvantage is the loss of information pertaining to the direction of blood flow, although this information can also be displayed.

Triphasic Normal Biphasic, low velocity, high resistance Distal obstruction

DOPPLER FLOW PATTERNS Normal Arteries The normal pattern of arterial blood flow in the extremity is different from that seen in the carotid arteries. At rest, the muscles of the extremities cause a high peripheral (distal) resistance and relatively low diastolic blood flow. The typical blood flow profile is a triphasic pattern (Fig. 26-2, top). First, during systole, there is a strong forward component of blood flow. Second, during early diastole, there is a short reversal of blood flow. Third, during remaining diastole, there is low-amplitude forward blood flow. The magnitude of the forward component of blood flow during diastole varies, disappearing with vasoconstriction caused be cold and increasing with warmth or after exercise (Fig. 26-3).

Stenotic Arteries The high-resistance pattern seen in normal peripheral arteries at rest is transformed into a low-resistance pattern when an occluded or severely stenotic arterial lesion is located proximal to the artery segment where the Doppler signals are sampled (see Fig. 26-2). This low-resistance pattern resembles that of the internal carotid artery. It is thought to reflect the opening of collateral arterial branches and the loss of normal resting arteriolar tone in response to ischemia. It is typically seen distal to an occluded artery segment but can be seen distal to severe stenotic lesions. A localized increase in velocity occurs at the site of a stenosis proper. This increase in blood flow velocity causes a shift in the Doppler frequency sampled at the stenosis. The Doppler frequency shift and increase in estimated flow velocity are proportional to the lumen diameter narrowing at the stenosis.10-12 This can be shown as an increase in color saturation or aliasing on the color Doppler map or as an increase in the peak systolic velocity on the Doppler spectral display (see Fig. 26-1, A and B). The pattern of blood flow distal to the stenosis is nonlaminar and shows a large variation in both direction and amplitude; this zone of disturbed flow is maintained over a distance of slightly more than 1 cm (see Video 26-1). In certain cases the zone of blood

Inflow obstruction

Monophasic, low velocity, lower resistance

Monophasic, high velocity, lower resistance Arteriovenous fistula

Biphasic, reciprocating Pseudoaneurysm

FIGURE 26-2.  Diagram of normal and abnormal Doppler arterial waveforms. The normal Doppler spectrum of flowing blood in the lower extremity arteries typically has a triphasic pattern: (1) forward flow during systole, (2) a short period of flow reversal in early diastole, and (3) lowvelocity flow during the remainder of diastole. Arterial Doppler signals are altered depending on the pathologic change. The four other patterns are examples of common arterial pathologies: distal obstruction, inflow obstruction, arteriovenous fistula, and pseudoaneurysm.

flow disturbance can be very small. This zone of disturbed blood flow is captured by the Doppler waveform as a broadening of the spectral window and by color Doppler imaging as increased variance of the color Doppler signals in the vessel.

Arteriovenous Fistulas Arteriovenous (AV) fistulas can be either congenital or iatrogenic. Congenital AV fistulas present in various forms as abnormal communications between an artery and large, distended venous channels or primary venous anomalies. The abnormalities more easily identified with Doppler ultrasound are usually quite obvious clinically and tend to be located close to the skin surface of the involved extremity.13 These are normally visualized as distended venous channels into which feed single or multiple arterial branches. Smaller, nondistended veins

Chapter 26  ■  The Peripheral Arteries   1001

Common femoral Profunda femoral Superficial femoral

Popliteal Tibio-peroneal trunk Anterior tibial Peroneal Posterior tibial

FIGURE 26-3.  Normal arterial waveforms. Doppler waveforms at the common femoral and popliteal arteries show triphasic patterns.

that have not dilated may still contain increased blood flow signals caused by the fistula. Iatrogenic communications often arise after selective arterial or venous catheterization or other forms of penetrating trauma. The communication can be visualized as a “jet of blood” (Video 26-2), with the involved vein distended compared to the other side. Blood flow signals in the recipient vein also show an arterial-like appearance, and the feeding artery can have increased diastolic blood flow (Fig. 26-4). The jet of blood has high-velocity signals, and on impact against the opposite vein wall, it can cause a perivascular vibration seen as an artifact on color Doppler imaging.14 An important differential diagnosis is compression of a vein by a hematoma. Venous compression causes a stenosis that increases blood flow velocity signals in the vein and mimics the high-velocity signals of a fistula (Fig. 26-5).

Masses The differential diagnosis of perivascular masses is facilitated by the use of color Doppler flow imaging, with some diagnostic specificity being offered by Doppler waveform analysis. Blood flow signals within a mass contiguous to an artery suggest the diagnosis of pseudoaneurysm. The communication tends to have a wide neck if the aneurysm arises at the anastomosis of a synthetic or autologous vein graft.15 With an iatrogenic pseudoaneurysm of the native artery, a small-diameter channel communicates to a larger, contained collection of blood. Color Doppler imaging shows blood flow

signals in the pseudoaneurysm cavity (Fig. 26-6; Video 26-3). A typical swirling motion or color yin-yang sign is typically seen within the collection itself.16,17 The Doppler waveform sampled in the communicating neck has a typical appearance (Fig. 26-6, A and B): the channel contains a backward-forward or a to-and-fro blood flow pattern.18 The to-and-fro pattern of blood flow shows rapid inflow into the cavity in systole and a slower, lower-amplitude exit of blood during diastole (Fig. 26-6, C and D). Hyperplastic lymph nodes and malignant lymph nodes can show both venous and arterial signals radiating from the hilum of the node (Fig. 26-7). These nodes can be mistaken for pseudoaneurysms.19,20 Points to consider in the differential diagnosis are (1) detection of arterial and venous signals where the communicating channel should be located and (2) absence of a to-andfro pattern of blood flow. Tumors can show a “rind” of hypervascularity at their periphery, as in thyroid gland adenomas. Arterial aneurysms are easily recognized by their typical location within the confines of the arterial wall. Although fusiform aneurysms follow this rule, it may be quite difficult to differentiate a saccular aneurysm from a pseudoaneurysm.21

PERIPHERAL ARTERY DISEASE Incidence and Clinical Importance Peripheral vascular disease is at least as prevalent as coronary artery disease or cerebrovascular disease.22

1002   PART III  ■  Small Parts, Carotid Artery, and Peripheral Vessel Sonography

Atherosclerosis is a generalized process in which the clinical presentation and development of symptoms depend on the arterial bed and the target organ. Coronary artery disease and carotid artery disease can present in a catastrophic manner, as myocardial infarction (MI) and cerebrovascular accident (CVA, stroke), respectively. This is very different from peripheral artery disease (PAD). Many patients have PAD for years before seeking medical assistance.23 This reflects the development of collateral arterial channels bypassing the diseased arterial segment as it progressively narrows. The collaterals are often sufficient to maintain perfusion to the lower extremity. The balance between blood supply and oxygen demand is maintained as long as the patient does not exercise or ambulate too vigorously. In general, patients with PAD can go on for years, decreasing their level of A

B FIGURE 26-4.  Arteriovenous (AV) fistula of femoral vessels after angiogram. A, Color flow Doppler image shows a high-velocity jet (arrow) from the common femoral artery (A) into the distended common femoral vein (V). B, The arterialtype signals sampled in the common femoral vein are consistent with a large AV fistula showing an arterialized venous blood flow pattern.

FIGURE 26-5.  Extrinsic compression of the common femoral vein (V) by a large hematoma (H) causes an increase in blood flow velocity. This can mimic the increased velocity seen in veins where an arteriovenous fistula is present.

RT PSA NECK

A

B

C

FIGURE 26-6.  Systolic and diastolic images of pseudoaneurysm. A, Image during end systole shows filling of the pseudoaneurysm and swirling motion of blood. B, Image at end systole shows emptying of the pseudoaneurysm through a small, communicating channel. C, Spectral Doppler tracing from the neck channel between the common femoral artery and the perivascular collection shows the classic to-and-fro waveform of a pseudoaneurysm.

Chapter 26  ■  The Peripheral Arteries   1003

A

B

FIGURE 26-7.  Groin lymph node with Doppler signal. A, Color Doppler signals in the soft tissues of the groin are complex. Careful examination shows that these signals are from the center of a structure that is a hyperplastic lymph node (arrows). A, Femoral artery; V, femoral vein. B, Spectral Doppler waveform from the center of this mass confirms the presence of a mainly arterial waveform and not the to-and-fro waveform of a pseudoaneurysm.

activity as their disease progresses. Disabling claudication is therefore more likely to be a presenting symptom in the younger patient with high levels of daily activity. The patient may also seek medical assistance because of the development of chronic changes of arterial insufficiency and poor wound healing. Acute embolic events originating from a more proximal arterial lesion, either from ulcerated plaques or popliteal aneurysms, can cause acute ischemia and extensive tissue loss, leading to amputation unless an intervention is performed. The widespread use of arterial bypass surgery has modified the natural history of PAD. The high patency rates of both arterial bypass surgery and similar patency rates for angioplasty allow patients who previously would have had amputation now to remain asymptomatic24,25 until other causes of mortality intercede. Acute cardiovascular events (e.g., MI, sudden death) are common causes of mortality in these patients, who already have generalized atherosclerosis.

Sonographic Technique Duplex Doppler sonography with gray-scale and Doppler spectral analysis is well accepted as the primary noninvasive modality for detecting evidence of lower extremity bypass graft dysfunction. It can also be used to evaluate the success of peripheral angioplasty, atherectomy, and stent placement.26-31 Doppler imaging of the leg arteries to determine the extent and nature of arterial lesions has become practical with the aid of color Doppler flow imaging. Although duplex Doppler sonography can be used to determine the presence of significant arterial lesions, the task of evaluating the whole leg is labor and time intensive. It takes 30 to 60 minutes to map out the

arterial tree of each leg using Duplex ultrasound.32 With color Doppler mapping, this task can be accomplished in 15 to 20 minutes.1 Color Doppler imaging also improves the accuracy of Doppler ultrasound as a diagnostic test for detecting and grading the severity of PAD.2,29

Lower Extremity Normal Anatomy The deep arteries of the leg travel with an accompanying vein. The common femoral artery starts at the level of the inguinal ligament and continues for 4 to 6 cm until it branches into the superficial and deep femoral arteries (see Fig. 26-3). The deep femoral artery quickly branches to supply the region of the femoral head and the deep muscles of the thigh. With PAD, collateral pathways often form between this deep femoral artery and the lower portions of the superficial femoral or the popliteal arteries. The superficial femoral artery continues along the medial aspect of the thigh at a depth of 4 to 8 cm until it reaches the adductor canal. At the boundary of the adductor canal, the superficial femoral artery continues as the popliteal artery. The popliteal artery crosses posterior to the knee, sending off small geniculate branches, and terminates as two major branches: the anterior tibial artery and tibioperoneal trunk. The anterior tibial artery courses in the anterior compartment of the lower leg after crossing through the interosseous membrane. It finally crosses the ankle joint as the dorsalis pedis artery. The tibioperoneal trunk gives off the posterior tibial and the peroneal arteries, which supply the calf muscles. The posterior tibial is more superficial than the peroneal artery and can be

1004   PART III  ■  Small Parts, Carotid Artery, and Peripheral Vessel Sonography

followed down to its typical location behind the medial malleolus. The blood flow pattern in all these branches is triphasic (Fig. 26-8; see also Fig. 26-2). There is an early systolic acceleration in velocity, followed by a brief period of low-amplitude flow reversal before returning to antegrade diastolic flow of low velocity. This pattern can be more pulsatile in the profunda femoris artery. Peak systolic velocity (PSV) varies with the level of the artery,

typically 100 cm/sec at the common femoral, down to 70 cm/sec at the popliteal artery. The tibioperoneal arteries have PSV of 40 to 50 cm/sec. The response to either exercise or transient ischemia is a loss of the triphasic pattern and the development of a monophasic pattern with antegrade blood flow with loss of early diastolic blood flow reversal (Fig. 26-9). Although a monophasic pattern can be seen in lower extremity disease or after exercise, PSV will be decreased in the ischemic limb of a patient with PAD, whereas it is increased in a healthy individual after exercise.

Aneurysms

FIGURE 26-8.  Normal triphasic waveform of lower extremity arteries.

Diagnostic Criteria.  Aneurysms develop as the structural integrity of the arterial wall weakens. Focal enlargement of the artery is more likely to occur at the level of the popliteal or distal superficial femoral artery (Fig. 26-10). Aneurysms are often bilateral and can remain asymptomatic for long periods. Ultrasound has become a gold standard in itself for confirming this suspected diagnosis.33,34 Although ultrasound can visualize the progressive thrombosis that fills in the aneurysm lumen to the level of the dilated wall, the lumen can appear normal at angiography. Ultrasound can be used to follow these aneurysms, as done for abdominal aneurysms. Unfortunately, no strict size criteria can be used to determine surgical suitability. Empirically, a 2-cm cutoff has been adopted.35 The development of symptoms suggestive of

A

Common femoral

V

Profunda femoral

Triphasic-proximal

Superficial femoral

Collateral

Collateral

Popliteal

V

Monophasic (low outflow distal resistance)

A

Monophasic (high outflow resistance)

FIGURE 26-9.  Significant arterial disease alters Doppler waveform. Sampling occurring distal to an occlusion. Doppler waveform sampled proximal to a high-grade stenosis may be normal or can show loss of the early and then late components of diastolic flow. Distal waveform is monophasic, most often with a relatively strong diastolic component; pattern is called a tardus-parvus waveform.

Chapter 26  ■  The Peripheral Arteries   1005

A

C

distal embolization by the thrombus accumulating in the lumen is an absolute indication for surgical intervention, regardless of the size of the aneurysm.35 Aneurysms will typically occlude with time due to accumulating thrombus (Fig. 26-10, B). Surgical exclusion (ligation of the aneurysm) is the traditional treatment. Doppler ultrasound can be used to monitor the success of the intervention.36 Aneurysm exclusion with covered stents is an alternative therapy to surgical intervention (Fig. 26-11). Doppler ultrasound can be used to monitor the patency of the stent and the exclusion of the aneurysm from the circulation37,38 (Video 26-4). Doppler techniques are useful in confirming the continued patency or occlusion of the lumen within the aneurysm. A bulge or focal enlargement of 20% of the expected vessel diameter constitutes a simple functional definition of an aneurysm. Serial monitoring should be considered in patients with small aneurysms less than 2 cm in size. Diagnostic Accuracy.  Direct pathologic verification of aneurysms diagnosed by ultrasound has shown that the technique is sensitive and specific and also superior to contrast angiography. The accuracy of Doppler techniques for confirming patency or occlusion of the lumen at the level of the aneurysm has yet to be reported, but it is accepted as a gold standard.

B

FIGURE 26-10.  Popliteal artery aneurysm. A, Transverse image of the popliteal space demonstrates a large aneurysm (arrows) with swirling blood flow pattern. B, Transverse image of the popliteal space of another patient shows a thrombosed popliteal artery (arrows) displacing the contiguous duplicated popliteal veins. C, Longitudinal gray-scale image of a different patient shows a fusiform aneurysm with a large amount of thrombus seen in the anterior and posterior walls.

Stenoses and Occlusions Diagnostic Criteria.  The effects of peripheral arterial lesions are detectable by a change in the blood flow pattern seen on the arterial Doppler waveform. At the lesion, peak systolic velocity increases (Fig. 26-12; see also Fig. 26-1 and Video 26-1), and early diastolic velocity reversal disappears. Distal to a moderately severe arterial lesion, the early diastolic blood flow reversal decreases and ultimately disappears as the lesion becomes more severe, and peak systolic blood flow velocity will decrease. The diastolic portion of the waveform increases in significance with respect to the decreasing peak systolic blood flow. On occasion, a high-resistance, monophasic pattern with absent diastolic blood flow can be seen, likely caused by peripheral vasoconstriction. The lowresistance pattern distal to the lesion is accentuated as the severity of the lesion increases. With severe lesions, the blood flow pattern is mainly that of forward flow, with end diastolic velocity approaching in amplitude the severely depressed peak systolic velocity. One explanation for the development of this pattern is progressive dilation of the arterioles within the distant vascular bed due to the release of metabolites caused by local ischemia. Another is the development of many small collateral branches that diminish the effective resistance

1006   PART III  ■  Small Parts, Carotid Artery, and Peripheral Vessel Sonography

IN STENT RT

B

A

Velocity

FIGURE 26-11.  Occlusion of popliteal artery stent. A, Color Doppler image of the proximal portion of a covered stent used to exclude a popliteal artery aneurysm. Blood flow is seen around the occluded stent, in effect causing a type 1 endoleak. B, Spectral Doppler tracing inside the stent confirming occlusion. Some faint venous signals from the contiguous popliteal vein are detected because of the high gain settings.

Time 2 to 4 cm proximal

of the distal arterial bed. This pattern is present in most cases of sufficiently severe proximal lesions, but it may not be seen when sampling within an artery segment proximal to tandem lesions such as distal high-grade focal lesions or occlusions. Signals in the artery proximal to a high-grade lesion can show a high-resistance pattern (see Fig. 26-9). With absent collaterals, forward blood flow can sometimes only be maintained during systole. The low-resistance pattern, a slow-rise low-amplitude pattern seen distal to segmental occlusions, is called the tardusparvus waveform (see Fig. 26-9). Although seen in most arterial segments distal to occlusions (Fig. 26-13), the

Velocity

FIGURE 26-12.  Blood flow velocity alterations occur with stenosis of at least 50%. Proximal to the lesion, the flow pattern is normal. At the stenosis, the peak systolic velocity increases in proportion to the degree of stenosis. Alterations in the diastolic portion of the Doppler waveform sampled at the lesion depend on the state of the distal arteries and the severity and geometry of the lesion; diastolic flow may increase dramatically or may be almost absent.

Time At the stenosis

low-resistance blood flow pattern may be absent when there is peripheral vasoconstriction. Color Doppler imaging can be used to survey the lower extremity arteries and identify likely stenotic lesions, through sites where color Doppler image shows aliasing (see Video 26-1). Focal areas where the measured PSV more than doubles from a contiguous and normal segment have been shown to correspond to lesions of greater than 50% narrowing in the lumen diameter of the artery.39 The velocity measured at the stenosis is divided by the velocity measured proximal to the stenosis. Peak systolic velocity is less sensitive to the

Chapter 26  ■  The Peripheral Arteries   1007

A

C

effects of vasodilation and vasoconstriction, so it is the preferred Doppler velocity parameter used to grade the severity of lower extremity arterial stenoses. Either end diastolic velocities (example EDVs; e.g., ≥80 cm/sec) or PSVs (>200 and >300 cm/sec) can be used as indicators of stenosis severity. However, EDV estimates are more variable than PSV measurements because EDV changes as a function of peripheral vasodilation. Diagnostic Accuracy and Applications.  In their original 1987 paper, Kohler et al.32 reported that Doppler sonography had a diagnostic sensitivity of 82% and a specificity of 92% for detecting segmental arterial lesions of the femoropopliteal arteries. They emphasized, however, that selective sampling had to be performed along the full course of the femoral and popliteal arteries. These segments normally measure 30 to 40 cm, so it is not surprising that such a survey took 1 to 2 hours to perform, especially if the iliac arteries were evaluated. Color Doppler sonography has been shown to reduce by 40% the time needed to examine the carotid artery for sites of suspected stenosis.9 A similar effect has been shown when color Doppler imaging is used to detect lower extremity arterial lesions.1 Diagnostic accuracy is

B

FIGURE 26-13.  Doppler waveforms above, at, and below occlusion of superficial femoral artery (SFA). A, Doppler image shows aliasing of color flow signals in the proximal SFA at the point of severe stenosis. Distally, loss of signal in the artery is caused by occlusion. Flow is present in area of stenosis because of collateral vessels (not shown). B, Doppler waveform sampled more distally confirms the absence of blood flow signals, indicating occlusion of the SFA (A). C, Downstream from the SFA occlusion, the popliteal artery signals are monophasic. Diastolic blood flow is low, probably because of peripheral vasoconstriction. There is blood flow in this popliteal artery from collateral blood supply (not shown).

also improved with color Doppler imaging compared with duplex sonography.2,40 With color Doppler the examination time is reduced to 30 minutes.1 Accuracy of color flow imaging of the peripheral arteries is almost 98% for distinguishing occlusions from nonoccluded segments. Accuracy for the detection of stenoses is greater than 85% for the femoropopliteal arteries,1,41-43 with some including an evaluation of the iliac arteries32,40 and runoff arteries.44 The evaluation of the runoff arteries is not as accurate as for the femoropopliteal system, especially for the peroneal artery.3,45,46 However, segments of the tibial arteries might be selected as suitable for the distal anastomosis of bypass grafts.47-49 Also, other imaging modalities might be unnecessary, relying exclusively on Doppler ultrasound before lower extremity bypass grafting.50-52 This is more likely for femoropopliteal bypass grafts.53 Color Doppler imaging is effective in triaging patients with symptoms of lower extremity arterial disease, reducing the need for diagnostic arteriography in more than half of patients presenting for clinical evaluation.54 Doppler sonography can also be used to triage patients likely to need peripheral angioplasty, therefore with

1008   PART III  ■  Small Parts, Carotid Artery, and Peripheral Vessel Sonography

A B

C

better management of more expensive imaging resources such as arteriography.29,55-57 No large studies have compared the efficacy of color Doppler imaging and CTA. However, cost-effectiveness analysis suggests that either color Doppler imaging or MRA can be used to select patients for interventions as a substitute for contrast arteriography.6 Color Doppler imaging and duplex sonography are extremely well suited for the evaluation of sites of percutaneous interventions such as angioplasty, atherectomy, and stent placement (Fig. 26-14). An original 1992 report indicated that one ultrasound measurement made a few days after angioplasty was predictive of lesion recurrence.58 Subsequent studies have failed to confirm this observation,59,60 but Doppler sonography can be used to detect recurrence of stenosis or occlusion at the site of a previous intervention. For example, postatherectomy results show a higher incidence of reocclusion than indicated by patients’ symptoms; atherectomy was not as efficient as angioplasty, with more lesion recurrences after atherectomy.61 Questions surround whether repeat imaging at the site of previous intervention is needed, because a repeat intervention might not be done if the patient remains asymptomatic.62 It does appear,

FIGURE 26-14.  Arterial stent. A, Transverse image shows an indwelling stent, seen as bright echoes (arrow) in the wall of the mid–superficial femoral artery. B, Spectral Doppler waveform in the stent is within normal limits at 84 cm/sec (arrow). C, Distal to the stent, the blood flow velocities are mildly elevated at 145 cm/sec (arrow).

however, that serial monitoring of sites of angioplasty and stent placement can predict technical success and lesion recurrence.63,64 No data indicate a benefit of re-intervention at the site of lesions detected by Doppler sonography.26 In fact, Doppler findings suggest that primary stent deployment is likely superior to stent deployment after angioplasty.30 Color Doppler imaging has been used to guide percutaneous interventions, angioplasty, and stent placement, without the use of contrast arteriography or fluoroscopy.59,65 Success rates have been high, but patient selection is critical to the success of the procedure.

Upper Extremity Normal Anatomy and Doppler Flow Patterns The arteries of the upper extremity are accompanied by veins: typically, only one vein at the level of the subclavian vein, occasionally duplicated at the level of the axillary veins, always duplicated at the level of the brachial veins and more distally. The junction of the subclavian artery with either the right brachiocephalic

Chapter 26  ■  The Peripheral Arteries   1009

(innominate) or the left brachiocephalic artery can be identified using an imaging window superior to the sternoclavicular joint. The artery is located superficial to the vein when the transducer is placed in the supraclavicular fossa. Near the junction of the middle and proximal thirds of the clavicle, it is necessary to use a window with the transducer placed on the chest, below the clavicle. The artery now lies deep to the subclavian vein. The origin of the axillary artery is lateral to the first rib, normally near the junction of the cephalic and the axillary vein. The axillary artery can be followed as it courses medially over the proximal humerus as it becomes the brachial artery. In most subjects the artery can be followed to the antecubital fossa, where it trifurcates into the radial, ulnar, and interosseous branches. The radial and ulnar branches can normally be imaged to the level

A

of the wrist. It is also possible to visualize the smaller digital branches. The normal flow pattern is triphasic and similar to the pattern seen in the leg.

Pathophysiology and Diagnostic Accuracy Most clinical interest in the noninvasive evaluation of the upper extremity arterial branches focuses on the (1) confirmation of pseudoaneurysms, (2) detection of focal stenosis caused by thoracic outlet syndrome (Fig. 26-15), (3) confirmation of native arterial occlusion secondary to emboli or trauma (Fig. 26-16), (4) detection of complications following cardiac catheterization, (5) evaluation of dialysis shunts, and (6) preoperative evaluation of radial artery patency.

B

FIGURE 26-15.  Thoracic outlet syndrome. A, Normal baseline subclavian artery waveform. B, Altered waveform during hyperextension, with compression against the clavicle causing stenosis and thoracic outlet syndrome.

A

B

FIGURE 26-16.  Normal and abnormal waveforms in brachial arteries. A, Normal waveform pf left side of brachial artery resembles the triphasic waveform seen in lower extremity arteries. B, Abnormal waveform of right side of brachial artery is obtained distal to a subclavian artery occlusion. The spectral Doppler waveform shows a low-amplitude (parvus) waveform with a slow systolic rise (tardus). This tardus-parvus waveform is typical distal to an arterial occlusion.

1010   PART III  ■  Small Parts, Carotid Artery, and Peripheral Vessel Sonography

Stenosis can be induced in the artery of patients with thoracic outlet syndrome by positioning the arm in the orientation that normally elicits symptoms, most often abducted (Fig. 26-15). Thoracic outlet syndrome is associated with distal arterial embolization, probably through mechanical forces predisposing the artery to develop an aneurysm. Thrombus then forms in the aneurysm and can embolize into the digital arteries. The extent of these acute or chronic occlusions must be mapped to assess the feasibility of bypass surgery before subjecting the patient to angiography. Proximal stenoses and occlusions associated with vasculitis can also be confirmed.66 After cardiac catheterization, suspected occlusions can be rapidly confirmed. Large hematomas can be readily evaluated and underlying pseudoaneurysms from jeopardized arteriotomy sutures confirmed or excluded. The radial artery is occasionally used as an access site for cardiac catheterization. Pseudoaneurysms can develop following cardiac catheterization67 (Fig. 26-17). The radial artery can also be harvested and serve as a donor conduit for coronary bypass surgery. Confirmation of the integrity of the palmar arch of the hand (dominant ulnar artery) is a prerequisite before harvest of the radial artery. This can be tested with Doppler ultrasound, imaging of the distal radial artery, and confirming reversal of blood flow on compression of the more proximal radial artery.68,69 Ulnar blood flow should increase when the radial artery is compressed and occluded.69

VASCULAR AND PERIVASCULAR MASSES Doppler sonography and color flow Doppler imaging have the ability to document the presence or absence of blood flow within masses located close to vessels or vascular prostheses. Although the presence of blood flow within a perivascular mass can be diagnostic of a pseudoaneurysm, the absence of blood flow makes it easier to justify a more conservative approach. In the case of a suspected hematoma, serial follow-up examinations can be used to document resolution of the process. In the

FIGURE 26-17.  Radial artery pseudoaneurysm. Pseudoaneurysm (arrows) arising in the radial artery after cardiac catheterization.

case of a suspected abscess, a biopsy can be performed without fear of uncontrolled hemorrhage.

Synthetic Vascular Bypass Grafts The various complications likely to affect the function of synthetic lower extremity bypass grafts15,70 are a function of the type of graft and time since placement (Fig. 26-18). In the first and second years after surgery, graft failure can result from technical errors or development of fibrointimal lesions at the anastomoses. Later failures may be caused by the progression of atherosclerotic lesions in the native vessels proximal and distal to the graft. The late complication of an anastomotic pseudo­ aneurysm occurs on average 5 to 10 years after graft placement and preferentially affects the femoral anastomosis of aortofemoral grafts.15,71 Infections can occur at any time after graft placement and may be associated with development of anastomotic pseudoaneurysm. With time, atherosclerotic changes and fibrointimal hyperplastic lesions mixed in with areas of chronic thrombus deposition can also develop in the synthetic graft conduit.

Masses: Hematoma versus Pseudoaneurysm Although the diagnostic accuracy of duplex Doppler sonography is greater than 95% for making the diagnosis of pseudoaneurysms at the anastomoses of bypass grafts, no specific waveform patterns have been described.72,73 The addition of color Doppler imaging can reveal an almost classic appearance of swirling motion of blood in the perivascular mass.15 This sign is not specific to a pseudoaneurysm because saccular aneurysms share similar flow patterns. The differential diagnosis is normally made when careful real-time imaging confirms that the mass is situated beyond the normal lumen of the vessel. The to-and-fro sign seen in native

FIGURE 26-18.  Synthetic graft. Gray-scale appearance of a synthetic PTFE lower extremity bypass graft (arrows).

Chapter 26  ■  The Peripheral Arteries   1011

pseudoaneurysms is obtained from Doppler spectral analysis of the signal sampled in the communicating channel between the perivascular collection and the native vessel. This neck often does not exist or is very broad, abutting the artery rather than extending as a thin structure for a length of a few centimeters. Typically, anastomotic pseudoaneurysms do not have any distinct communicating channels. Care must be taken to differentiate perivascular pulsations transmitted within a hematoma from flowing blood. Adjustment of the flow sensitivity of the imaging device to minimize this artifact in the normal artery proximal or distal to the site of abnormality can help eliminate this error. Setting the color velocity scale (peak repetition frequency, PRF) to a high value can eliminate this artifact while not hindering the detection of the communicating channel.

Occlusions and Anastomotic Stenoses The absence of Doppler signals within a bypass graft is diagnostic of an occlusion. An anastomotic stenosis will typically cause a marked increase in the Doppler velocity signals sampled at the anastomosis or beyond. Normally, however, blood flow velocities tend to increase as the graft tapers to the anastomosis. Increases in velocity caused by the geometry of the anastomotic connection are common and can cause up to a 100% increase in velocity without being indicative of a pathologic lesion. No studies have addressed the actual incidence and significance of this finding. Serial monitoring of these sites of disturbed flow may be done on the basis that an increase in velocity over a few months is indicative of a developing stenosis.74

AUTOLOGOUS VEIN GRAFTS Two types of venous bypass grafts are currently used for arterial revascularization: the reversed vein and the “in situ” vein grafts. The reversed vein is a segment of native superficial vein that has been harvested from its normal anatomic location, reversed, and then anastomosed to the native artery segments proximal and distal to the diseased segments. The in situ technique typically uses the greater saphenous vein, although the lesser saphenous vein can be used for popliteal-to-distal tibioperoneal bypass surgery. The vein is left in its native bed. The valves are lysed and the side branches (perforating veins that normally communicate to deep venous system) are ligated. The proximal and distal portions are mobilized and anastomosed to the selected arterial segments. Three different mechanisms are responsible for bypass graft failure. Early failures are seen within 1 month of surgery and usually result from technical errors, including poor suture line placement, opening of unsuspected

venous channels in the in situ grafts, poor selection of anastomotic sites, and poorly lysed or disrupted vein valves. For 2 years after surgery, fibrointimal or fibrotic lesions tend to develop either at the anastomosis or within the graft conduit, most often at the site of a vein valve. Later failures, after 2 years, are thought to result from the progressive atherosclerotic process in the native vessels proximal and distal to the anastomosis.

Stenosis of Venous Bypass Graft A decreased blood flow velocity within a vein bypass graft indicates a high likelihood of incipient graft occlusion and thrombosis (Fig. 26-19). Bandyk et al.75,76 have shown that PSV less than 40 or 45 cm/sec can be used to identify such grafts. This diagnostic criterion can appropriately identify only the more severely diseased grafts.77 It does not identify the site of stenoses likely to progress until they become flow restrictive and finally result in graft thrombosis.78 The lesions that develop within bypass grafts are most often the result of fibrointimal hyperplasia, and their presence must be identified before they can be monitored for possible progression of severity. Color Doppler sonography can be used to survey the 30 to 80 cm–long bypass graft very efficiently. The site of a suspected stenosis can be quickly identified and Doppler spectral analysis used to grade the severity of the stenosis using the PSV ratio (Fig. 26-20). Power Doppler imaging and “B-flow imaging” (a technique that visualizes moving blood) can also be used to better confirm the presence of any stenotic lesions. The PSV ratio is calculated by dividing the peak-systolic velocity measured at the suspected stenosis by that measured in the portion of the graft 2 to 4 cm proximal (Fig. 26-21). Blood flow velocity ratios of 2 or more correspond to 50% diameter stenosis.32,57 Blood flow velocity ratios of 3 or more correspond to 75% diameter stenosis.27,57 Critical stenoses have been empirically identified as those causing a velocity increase by a factor of 3.5, 3.7,

FIGURE 26-19.  Abnormal flow velocity of bypass graft. Depressed velocity (