Gynecologic Imaging E-Book
1058 pages

Vous pourrez modifier la taille du texte de cet ouvrage

Gynecologic Imaging E-Book


Obtenez un accès à la bibliothèque pour le consulter en ligne
En savoir plus
1058 pages

Vous pourrez modifier la taille du texte de cet ouvrage

Obtenez un accès à la bibliothèque pour le consulter en ligne
En savoir plus


Gynecologic Imaging, a title in the Expert Radiology Series, by Drs. Julia R. Fielding, Douglas Brown, and Amy Thurmond, provides the advanced insights you need to make the most effective use of the latest gynecologic imaging approaches and to accurately interpret the findings for even your toughest cases. Its evidence-based, guideline-driven approach thoroughly covers normal and variant anatomy, pelvic pain, abnormal bleeding, infertility, first-trimester pregnancy complications, post-partum complications, characterization of the adnexal mass, gynecologic cancer, and many other critical topics. Combining an image-rich, easy-to-use format with the greater depth that experienced practitioners need, it provides richly illustrated, advanced guidance to help you overcome the full range of diagnostic, therapeutic, and interventional challenges in gynecologic imaging. Online access at allows you to rapidly search for images and quickly locate the answers to any questions.

  • Get all you need to know about the latest advancements and topics in gynecologic imaging, including normal and variant anatomy, pelvic pain, abnormal bleeding, infertility, first-trimester pregnancy complications, post-partum complications, characterization of the adnexal mass, and gynecologic cancer.
  • Recognize the characteristic presentation of each disease via any modality and understand the clinical implications of your findings.
  • Consult with the best. Internationally respected radiologist Dr. Julia Fielding leads a team of accomplished specialists who provide you with today’s most dependable answers on every topic in gynecologic imaging.
  • Identify pathology more easily with 1300 detailed images of both radiographic images and cutting-edge modalities—MR, CT, US, and interventional procedures.
  • Find information quickly and easily thanks to a consistent, highly templated, and abundantly illustrated chapter format.

Access the fully searchable text online at, along with downloadable images.


United States of America
Placenta previa
Ovarian pregnancy
Intrauterine device
Cervical pregnancy
Urge incontinence
Hormone replacement therapy
The Only Son
Chorioadenoma destruens
Fallopian tube obstruction
Vaginal intraepithelial neoplasia
Postpartum hemorrhage
Pelvic pain
Contrast medium
Female infertility
Diffusion MRI
Iodinated contrast
Gestational trophoblastic disease
Hydatidiform mole
Rectovaginal fistula
Chronic kidney disease
Acute kidney injury
Pulmonary hypertension
Abdominal pain
Deep vein thrombosis
Physician assistant
Uterine cancer
Pancreatic cancer
Ovarian cyst
Bowel obstruction
Follicle-stimulating hormone
Health care
Medical imaging
Obstetric fistula
Pulmonary embolism
Fecal incontinence
Urinary incontinence
Medical ultrasonography
Tubal ligation
In vitro fertilisation
Urinary system
Insulin resistance
Ectopic pregnancy
Polycystic ovary syndrome
Obstetrics and gynaecology
X-ray computed tomography
Turner syndrome
Kidney stone
Varicose veins
Data storage device
Radiation therapy
Pelvic inflammatory disease
Positron emission tomography
Magnetic resonance imaging


Publié par
Date de parution 05 avril 2011
Nombre de lectures 1
EAN13 9781437735987
Langue English
Poids de l'ouvrage 6 Mo

Informations légales : prix de location à la page 0,0770€. Cette information est donnée uniquement à titre indicatif conformément à la législation en vigueur.


Gynecologic Imaging

Julia R. Fielding, MD
Professor of Radiology, Division Chief of Abdominal Imaging, Department of Radiology, University of North Carolina at Chapel Hill, Chapel Hill, North Carolina

Douglas L. Brown, MD
Professor of Radiology, Department of Radiology, Mayo Clinic College of Medicine, Rochester, Minnesota

Amy S. Thurmond, MD
Director, Medical Women’s Imaging Department, Siker Medical Imaging and Intervention of Portland, Portland, Oregon
Front Matter

Gynecologic Imaging
Julia R. Fielding, MD
Professor of Radiology, Division Chief of Abdominal Imaging, Department of Radiology, University of North Carolina at Chapel Hill, Chapel Hill, North Carolina
Douglas L. Brown, MD
Professor of Radiology, Department of Radiology, Mayo Clinic College of Medicine, Rochester, Minnesota
Amy S. Thurmond, MD
Director, Medical Women's Imaging Department, Siker Medical Imaging and Intervention of Portland, Portland, Oregon

1600 John F. Kennedy Boulevard
Suite 1800
Philadelphia, PA 19103-2899
Copyright © 2011 Saunders, an imprint of Elsevier Inc.
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: .
This book and the individual contributions contained in it are protected under copyright by the Publisher (other than as may be noted herein).

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.
Library of Congress Cataloging-in-Publication Data
Gynecologic imaging / [edited by] Julia Fielding, Douglas Brown, Amy Thurmond.
p. ; cm. – (Expert radiology series)
Includes bibliographical references.
ISBN 978-1-4377-1575-0 (hardcover)
1. Generative organs, Female--Imaging. 2. Generative organs, Female–Diseases–Diagnosis. 3. Pregnancy–Complications–Diagnosis. I. Fielding, Julia R. II. Brown, Douglas (Douglas L.) III. Thurmond, Amy S. IV. Series: Expert radiology series.
[DNLM: 1. Genitalia, Female–radiography. 2. Diagnostic Imaging–methods. 3. Genital Diseases, Female–diagnosis. 4. Pregnancy Complications–radiography. WP 141]
RG107.5.I4G96 2011
618.1'075–dc22 2011005048
Acquisitions Editor: Pamela Hetherington
Developmental Editor: Kristina Oberle
Publishing Services Manager: Anne Altepeter
Project Managers: Beth Hayes and Jessica L. Becher
Design Direction: Steven Stave
Printed in China
Last digit is the print number: 9 8 7 6 5 4 3 2 1

Lejla Aganovic, MD , Assistant Professor, Department of Radiology, University of California, San Diego, California
Computed Tomography: Normal Anatomy, Imaging Techniques, and Pitfalls

Paula Amato, MD , Associate Professor, Department of Obstetrics and Gynecology, Oregon Health and Science University, Portland, Oregon
Infertility, Evaluation, and Treatment

Rochelle F. Andreotti, MD , Professor of Clinical Radiology and Radiology Sciences, Department of Radiology, Associate Professor of Clinical Obstetrics and Gynecology, Department of Obstetrics and Gynecology, Vanderbilt University Medical Center, Nashville, Tennessee
Approach to Pelvic Pain and the Role of Imaging

Mostafa Atri, MD, Dipl Epid , Head, Section of Ultrasound, Joint Department of Medical Imaging, Head, Division of the Abdomen, Medical Imaging Department, Toronto University, Professor of Radiology, University of Toronto, Toronto, Ontario, Canada
Malignant Ovarian Masses

Deborah A. Baumgarten, MD, MPH , Associate Professor, Department of Radiology, Emory University, Atlanta, Georgia
Benign Endometrial Causes of Abnormal Bleeding

Douglas L. Brown, MD , Professor of Radiology, Department of Radiology, Mayo Clinic College of Medicine, Rochester, Minnesota
Pelvic Pain: Lower Urinary Tract—Urethral Diverticulum, Cysts, and Varix
Uterine Leiomyomas
Approach to Imaging the Adnexal Mass

Lauren M. Brubaker, MD , Resident Physician, Department of Radiology, University of North Carolina Hospitals, University of North Carolina at Chapel Hill, Chapel Hill, North Caroliina
Hysterosalpingography: Techniques, Normal Anatomy, and Pitfalls

Lara Lyn Bryan-Rest, MD , Department of Diagnostic Radiology, Yale–New Haven Hospital, New Haven, Connecticut
Ectopic Pregnancy

Richard L. Clark, MD , Emeritus Professor of Radiology, Department of Radiology, University of North Carolina School of Medicine, Chapel Hill, North Carolina
Hysterosalpingography: Techniques, Normal Anatomy, and Pitfalls

Harris L. Cohen, MD , Chair and Professor of Radiology, Pediatrics, and Obstetrics and Gynecology, Department of Radiology, University of Tennessee, Medical Director, Department of Radiology, LeBonheur Children's Medical Center, Memphis, Tennessee, Emeritus Professor of Radiology, Stony Brook School of Medicine, Stony Brook, New York
Gynecologic Imaging of the Pediatric Patient

Carlos Cuevas, MD , Assistant Professor, Department of Radiology, University of Washington, Director, Gastronterology Radiology, Department of Radiology, University of Washington Medical Center, Seattle, Washington
Chronic Pelvic Pain

, MD Roberta diFlorio-Alexander, Assistant Professor, Radiology and Obstetrics and Gynecology, Dartmouth Medical School, Hanover, New Hampshire
Postpartum Complications

Manjiri Dighe, MD , Assistant Professor, Department of Radiology, University of Washington Medical Center, Seattle, Washington
Acute Pelvic Pain
Chronic Pelvic Pain

Vikram Dogra, MD , Professor of Radiology, Urology and BME, Department of Imaging Science, University of Rochester Medical Center, Rochester, New York
The Normal Pelvis on Ultrasound Imaging and Anatomic Correlations

Peter M. Doubilet, MD, PhD , Professor of Radiology, Harvard Medical School, Senior Vice Chair of Radiology, Brigham and Women's Hospital, Boston, Massachusetts
Ultrasound-Guided Treatment of Ectopic Pregnancy

Theodore Dubinsky, MD, FSRU , Larry Mack Professor of Radiology, Obstetrics, Gynecology, and Reproductive Health Sciences, Director of Body Imaging, Department of Radiology, University of Washington School of Medicine, Seattle, Washington
Chronic Pelvic Pain

Sara Durfee, MD , Assistant Professor, Department of Radiology, Harvard Medical School, Associate Radiologist, Department of Radiology, Brigham and Women's Hospital, Boston, Massachusetts
Retained Products of Conception

Steven C. Eberhardt, MD , Associate Professor, Department of Radiology, University of New Mexico, Albuquerque, New Mexico
Ovarian and Fallopian Tube Cancer

Fiona M. Fennessy, MD, PhD , Assistant Professor of Radiology, Department of Radiology, Harvard Medical School, Assistant Professor of Radiology, Department of Radiology, Brigham and Women's Hospital, Boston, Massachusetts
Magnetic Resonance–Guided Ultrasound Surgery of Uterine Leiomyomas

Julia R. Fielding, MD , Professor of Radiology, Division Chief of Abdominal Imaging, Department of Radiology, University of North Carolina at Chapel Hill, Chapel Hill, North Carolina
Magnetic Resonance Imaging of the Female Pelvis: Technique, Anatomy, and Pitfalls
Imaging of Pelvic Floor Dysfunction

Maureen S. Filipek, MD , Radiologist, Women's Imaging, EPIC Imaging West, Beaverton, Oregon
Gestational Trophoblastic Neoplasia
Uterine Cancers

Jurgen J. Fütterer, MD, PhD , Radiologist, Department of Radiology, Radboud University Nijmegen Medical Centre, Nijmegen, The Netherlands
Diffusion Magnetic Resonance Imaging

Margaret L. Gallegos, BS, MD , Resident, Department of Radiology, University of New Mexico, Albuquerque, New Mexico
Ovarian and Fallopian Tube Cancer

Nancy Hammond, MD , Assistant Professor of Radiology, Department of Radiology, Feinberg School of Medicine, Northwestern University, Chicago, Illinois
Uterine Artery Embolization

Robert D. Harris, MD, MPH , Co-Director of Ultrasound (Education and Research), Department of Radiology, Dartmouth Medical School, Lebanon, New Hampshire, Professor of Radiology and Obstetrics and Gynecology, Dartmouth-Hitchcock Medical Center, Hannover, New Hampshire
Postpartum Complications

David S. Hartman, MD , Professor of Radiology, Department of Radiology, The Pennsylvania State College of Medicine, Professor of Radiology, Department of Radiology, Milton S. Hershey Medical Center, Hershey, Pennsylvania
The Imaging of Contraception

Sara M. Harvey, MD , Instructor, Department of Radiology and Radiological Sciences, Vanderbilt University, Nashville, Tennessee
Approach to Pelvic Pain and the Role of Imaging

Tara Henrichsen, MD , Instructor, Department of Diagnostic Radiology, Mayo Clinic, Rochester, Minnesota
Approach to Imaging the Adnexal Mass

Mindy M. Horrow, MD, FACR, FSRU, FAIUM , Associate Professor of Radiology, Department of Radiology, Thomas Jefferson University School of Medicine, Director of Body Imaging, Department of Radiology, Albert Einstein Medical Center, Philadelphia, Pennsylvania
Pitfalls in Gynecologic Ultrasound

Keyanoosh Hosseinzadeh, MD , Assistant Professor, Department of Diagnostic Imaging, University of Pittsburgh Medical Center, Section Chief, Body Magnetic Resonance Imaging, Department of Diagnostic Imaging, University of Pittsburgh Medical Center, Pittsburgh, Pennsylvania
Müllerian Uterine Anomalies

Golbahar Houshmand, MD , Research fellow, Department of Radiology, University of Pittsburgh Medical Center, Pittsburgh, Pennsylvania
Müllerian Uterine Anomalies

Lynne M. Hurwitz, MD , Associate Professor, Department of Radiology, Duke University Medical Center, Durham, North Carolina
Dose Reduction Techniques in Multidetector Computed Tomography Body Imaging

Tracy A. Jaffe, MD , Associate Professor of Radiology, Department of Radiology, Duke University Medical Center, Durham, North Carolina
Dose Reduction Techniques in Multidetector Computed Tomography Body Imaging

Keith C. Kaplan, MD, MS , Resident Physician, Department of Radiology, The Pennsylvania State College of Medicine, Resident Physician, Department of Radiology, Milton S. Hershey Medical Center, Hershey, Pennsylvania
The Imaging of Contraception

Akira Kawashima, MD, PhD , Professor, Department of Radiology, Mayo Clinic College of Medicine, Consultant, Department of Radiology, Mayo Clinic, Rochester, Minnesota
Pelvic Pain: Lower Urinary Tract—Urethral Diverticulum, Cysts, and Varix

Cheryl L. Kirby, BA, MD , Assistant Professor, Department of Radiology, Albert Einstein Medical Center, Philadelphia, Pennsylvania
Pitfalls in Gynecologic Ultrasound

Jill E. Langer, MD , Associate Professor, Department of Radiology, University of Pennsylvania School of Medicine, Hospital of the University of Pennsylvania, Philadelphia, Pennsylvania
Benign Ovarian Masses

Yan Mee Law, MBBS, FRCR , Department of Diagnostic Radiology, Singapore General Hospital, Singapore
Imaging of Pelvic Floor Dysfunction
Cervical Cancer

Ellie R. Lee, MD , Assistant Professor, Department of Radiology, Univeristy of North Carolina at Chapel Hill, Chapel Hill, North Carolina
Tubal Abnormalities

Susanna I. Lee, MD, PhD , Assistant Professor, Department of Radiology, Harvard Medical School, Chief of Women's Imaging, Department of Radiology, Massachusetts General Hospital, Boston, Massachusetts
Drainage and Biopsy Procedures

Thomas Lemond, BA, MD , Radiology Resident, Postgraduate Year 2, University of Tennessee Health Sciences Center, Memphis, Tennessee
Gynecologic Imaging of the Pediatric Patient

Andrew J. LeRoy, MD , Consultant, Department of Radiology, Mayo Clinic, Professor of Radiology, Mayo Clinic College of Medicine, Rochester, Minnesota
Pelvic Pain: Lower Urinary Tract—Urethral Diverticulum, Cysts, and Varix

Alfred Llave, MD , Abdominal Imaging Fellow, Radiology, University of North Carolina, Chapel Hill, North Carolina
Magnetic Resonance Imaging of the Female Pelvis: Technique, Anatomy, and Pitfalls

Shirley M. McCarthy, MD, PhD , Professor, Department of Diagnostic Radiology and Obstetrics and Gynecology, Yale University, New Haven, Connecticut

Sean E. McSweeney, FFRCSI, MB, BCh, BAO , Joint Department of Medical Imaging, University of Toronto, Toronto, Ontario, Canada
Malignant Ovarian Masses

Rashmi T. Nair, Cleveland Clinic Foundation, Cleveland, Ohio
Use of Positron Emission Tomography Imaging in Gynecologic Cancers

Paul Nikolaidis, MD , Associate Professor of Radiology, Feinberg School of Medicine, Northwestern University, Chicago, Illinois
Uterine Artery Embolization

Ifeyinwa Y. Onyiuke, MB, BS , Clinical Professor, Yale University School of Medicine, Attending Radiologist, Veterans Administration Connecticut Health Care System, New Haven, Connecticut

Tulin Ozcan, MD , Associate Professor, Department of Obstetrics and Gynecology, University of Rochester, Rochester, New York
The Normal Pelvis on Ultrasound Imaging and Anatomic Correlations

Raj Mohan Paspulati, MD , Assistant Professor, Department of Radiology, Case Western Reserve University, University Hospitals, Cleveland, Ohio
Carcinoma of the Vagina and Vulva

Philip E. Patton, MD , Professor, Department of Obstetrics and Gynecology, Oregon Health and Science University, Portland, Oregon
Infertility, Evaluation, and Treatment

Hope E. Peters, MD , Instructor of Radiology, Department of Ultrasound, Brigham and Women's Hospital, Boston, Massachusetts
Ultrasound-Guided Treatment of Ectopic Pregnancy

Misty Blanchette Porter, MD , Associate Professor, Departments of Obstetrics and Gynecologyand Radiology, Division of Reproductive Medicine and Infertility, Dartmouth Medical School, Hanover, New Hampshire, Medical Director, In Vitro Fertilization and Assisted Reproduction Technology, Departments of Obstetrics and Gynecologyand Radiology, Division of Reproductive Medicine and Infertility, Dartmouth-Hitchcock Medical Center, Lebanon, New Hampshire
The Ovary and Polycystic Ovary Syndrome

Dmitry Rakita, MD , Chief, Body Magnetic Resonance Imaging, Department of Radiology, Baystate Medical Center, Springfield, Massachusetts, Assistant Professor, Tufts University School of Medicine, Boston, Massachusetts

Leslie M. Scoutt, MD , Professor of Diagnostic Radiology and Surgery, Department of Diagnostic Radiology, Yale University School of Medicine, Chief, Ultrasound Service, Medical Director of the Non-Invasive Vascular Laboratory, Department of Diagnostic Radiology, Yale-New Haven Hospital, New Haven, Connecticut
Ectopic Pregnancy

Shetal N. Shah, MD , Assistant Professor, Department of Abdominal Imaging and Nuclear Medicine, Co-Director, Center for Positron Emission Technology and Molecular Imaging, Imaging Institute, Cleveland Clinic, Cleveland, Ohio
Use of Positron Emission Tomography Imaging in Gynecologic Cancers

Clare M. Tempany, MD , Professor, Department of Radiology, Harvard Medical School, Department of Radiology, Brigham and Women's Hospital, Boston, Massachusetts
Magnetic Resonance–Guided Ultrasound Surgery of Uterine Leiomyomas

Ashraf Thabet, MD , Instructor in Radiology, Department of Radiology, Harvard Medical School, Assistant Radiologist, Department of Radiology, Massachusetts General Hospital, Boston, Massachusetts
Drainage and Biopsy Procedures

Amy S. Thurmond, MD , Director, Medical Women's Imaging Department, Siker Medical Imaging and Intervention of Portland, Portland, Oregon
Vaginal Fistulas
Imaging of Pelvic Floor Dysfunction
Fallopian Tube Catheterization

Fauzia Q. Vandermeer, MD , Assistant Professor, Department of Diagnostic Radiology, University of Maryland School of Medicine, Baltimore, Maryland
Ultrasound of the Normal and Failed First-Trimester Pregnancy

Geoffrey E. Wile, BS, MD , Assistant Professor, Department of Radiology, Vanderbilt University Medical Center, Nashville, Tennessee
Approach to Pelvic Pain and the Role of Imaging

Jade Wong-You-Cheong, MBChB, FRCR , Professor, Department of Diagnostic Radiology, University of Maryland School of Medicine, Baltimore, Maryland
Ultrasound of the Normal and Failed First-Trimester Pregnancy
When Elsevier approached me with the idea of a textbook of gynecologic imaging I was ecstatic that a publisher was interested in promoting women's health care and a little worried that there just was not a need for another textbook. Then I started looking carefully at the reading rooms where the on-call radiologists work. Virtually every night two textbooks were open, one devoted to ultrasound and the other to computed tomography (CT), both with images depicting gynecologic pathology. Imaging of the female pelvis remains challenging because of the anatomic complexity; the varied appearance of gynecologic diseases on CT, ultrasound (US) and magnetic resonance imaging (MRI); and, often, the urgency of the clinical situation. I contacted my co-editors, Drs. Brown and Thurmond, and we began to plan.
Organization of Gynecologic Imaging is based on the questions and comments we have received during our course lectures, board review sessions, literature assessment, and interactions with both practicing radiologists and trainees. Our textbook is divided into sections reviewing CT, US, and MRI techniques; normal anatomy; and disease states. Particular attention is devoted to basic protocols that will provide the most critical clinical information. This includes Doppler imaging, CT contrast administration, and MR pulse sequences. Some types of benign disease of the pelvis have distinctive imaging features and thus are presented separately from malignant disease. Topics include infection, infertility, uterine masses, and pelvic floor dysfunction. We have also included images of disease states beyond the gynecologic tract that may mimic expected clinical findings. Cancers involving the uterus, cervix, ovaries, fallopian tubes, and vulva are explored, with clear depictions of disease, including line drawings and color prints of clinical findings. We have included the most current TMN staging system for each cancer adjacent to corresponding images. Chapters are also devoted to new and evolving techniques in the diagnosis of cancer, including positron emission tomography (PET) and MR diffusion imaging both during and after therapy. Current American College of Radiology Appropriateness Criteria are presented in an appendix.
As radiology has expanded to become the primary diagnostic and staging test for many gynecologic diseases, so has its role in therapy. Uterine artery emobilization and focused ultrasound of fibroids, abscess drainage, and fallopian tube canalization are presented, with expert advice on patient preparation, technique, and communication with referring physicians.
Ultrasound is particularly challenging because of the transfer of real-time observations to images. With physicians performing less hands-on scanning, we risk losing our ability to confidently identify common diseases and especially the complications of early pregnancy. Several chapters in this book are devoted to pelvic ultrasound, with numerous high-quality images, including both grayscale and color Doppler.
Finally, we have included a chapter specifically devoted to radiation safety. We all want to decrease dose to our patients, but not at the expense of a diagnostic test. The authors of this chapter discuss specific steps to decrease dose and provide clinical scenarios and algorithms.
This book was written primarily for radiologists, gynecologists, and nurse practitioners who take care of women with pelvic disease. It provides a practical review of anatomy, appropriate imaging of benign and malignant disease of the female pelvis, and guidelines for patient management. In editing this book, we learned a great deal about our specialty. It is our hope that our readers will as well.

Julia R. Fielding

Douglas L. Brown

Amy S. Thurmond
To my parents, husband, and son who tolerated long days, late nights, and travels away from home with good humor; I owe my career to them.
To my mentors and colleagues, especially at Boston University Medical Center, Bringham and Women's Hospital, and the University of North Carolina at Chapel Hill; they continue to motivate me to strive for excellence
To my many residents and fellows who inspire me with their confidence and abilities; they give me great hope for the future of radiology and health care in America
To my co-editors, Dr. Amy Thurmond and Dr. Douglas Brown; all my thanks for your diligent work and patience

Julia R. Fielding
To Tina, for her continuing love and support, which I treasure
To the rest of my family, for their caring
To the many others who have also been a part of my journey, especially those at the University of Tennessee in Memphis, Brigham and Women's Hospital in Boston, and Mayo Clinic in Rochester – my mentors and colleagues who guided me; the sonographers, students, nurses, and support staff who helped me; the residents and fellows who motivated me; and the patients and their families who are the reason for what we do

Douglas L. Brown
To my ancestors from whom I inherited curiosity, drive, and the desire to make things better
To my immediate family, husband, and four children, who put up with the related qualities: stubbornness and a chaotic schedule. They've helped me so much.

Amy S. Thurmond
Table of Contents
Instructions for online access
Front Matter
Section One: Imaging Techniques, Pitfalls, and Normal Anatomy
Part One: Ultrasound
Chapter 1: The Normal Pelvis on Ultrasound Imaging and Anatomic Correlations
Chapter 2: Pitfalls in Gynecologic Ultrasound
Part Two: Computed Tomography
Chapter 3: Computed Tomography: Normal Anatomy, Imaging Techniques, and Pitfalls
Chapter 4: Dose Reduction Techniques in Multidetector Computed Tomography Body Imaging
Part Three: Magnetic Resonance Imaging
Chapter 5: Magnetic Resonance Imaging of the Female Pelvis: Technique, Anatomy, and Pitfalls
Chapter 6: Diffusion Magnetic Resonance Imaging
Part Four: Fluoroscopy
Chapter 7: Hysterosalpingography: Techniques, Normal Anatomy, and Pitfalls
Section Two: Pelvic Pain
Chapter 8: Approach to Pelvic Pain and the Role of Imaging
Chapter 9: Endometriosis
Chapter 10: Acute Pelvic Pain
Chapter 11: Chronic Pelvic Pain
Chapter 12: Pelvic Pain: Lower Urinary Tract—Urethral Diverticulum, Cysts, and Varix
Section Three: Abnormal Bleeding
Chapter 13: Benign Endometrial Causes of Abnormal Bleeding
Chapter 14: Adenomyosis
Chapter 15: Uterine Leiomyomas
Section Four: Infertility
Chapter 16: Infertility, Evaluation, and Treatment
Chapter 17: Tubal Abnormalities
Chapter 18: Müllerian Uterine Anomalies
Chapter 19: The Ovary and Polycystic Ovary Syndrome
Chapter 20: The Imaging of Contraception
Section Five: First-Trimester Pregnancy Complications
Chapter 21: Ultrasound of the Normal and Failed First-Trimester Pregnancy
Chapter 22: Ectopic Pregnancy
Chapter 23: Retained Products of Conception
Chapter 24: Gestational Trophoblastic Neoplasia
Section Six: Postpartum Complications
Chapter 25: Postpartum Complications
Chapter 26: Vaginal Fistulas
Chapter 27: Imaging of Pelvic Floor Dysfunction
Section Seven: Characterization of the Adnexal Mass
Chapter 28: Approach to Imaging the Adnexal Mass
Chapter 29: Benign Ovarian Masses
Chapter 30: Malignant Ovarian Masses
Section Eight: Gynecologic Cancer
Chapter 31: Use of Positron Emission Tomography Imaging in Gynecologic Cancers
Chapter 32: Uterine Cancers
Chapter 33: Cervical Cancer
Chapter 34: Ovarian and Fallopian Tube Cancer
Chapter 35: Carcinoma of the Vagina and Vulva
Section Nine: Pediatric Imaging
Chapter 36: Gynecologic Imaging of the Pediatric Patient
Section Ten: Interventional Radiology in Gynecology
Chapter 37: Drainage and Biopsy Procedures
Chapter 38: Magnetic Resonance–Guided Ultrasound Surgery of Uterine Leiomyomas
Chapter 39: Uterine Artery Embolization
Chapter 40: Fallopian Tube Catheterization
Chapter 41: Ultrasound-Guided Treatment of Ectopic Pregnancy
Selections from the ACR Appropriateness Criteria
Section One
Imaging Techniques, Pitfalls, and Normal Anatomy
Part One
Chapter 1 The Normal Pelvis on Ultrasound Imaging and Anatomic Correlations

Tulin Ozcan, Vikram Dogra
Gynecologic imaging by ultrasound (US) has improved significantly during the past 20 years with the advent of various scanning techniques. The incorporation of transvaginal ultrasound (TVUS) has become a routine part of the gynecologic evaluation. Compared with transabdominal ultrasound (TAUS), tissue attenuation is usually less of an issue with TVUS because it allows the probe to be placed closer to the pelvic tissue, and higher frequency probes can be used. Recent advances in three-dimensional (3D) imaging have allowed the visualization of the coronal uterine plane and further improved the diagnostic accuracy of TVUS imaging.

Technical Requirements
The quality of a pelvic US examination is dictated by the correct selection of probes and the scanning experience of the sonologist or sonographer. The transducers and probes are characterized by their scanning area and their frequency. The scanning area is seen as rectangular with linear probes, whereas it is triangular with sector probes. The convex (curvilinear) probes are used most commonly, and the footprint will depend on their curvature. As the frequency of the probe is increased, the beam wavelength shortens and the resolution improves. The penetration decreases, however, with increasing frequency of the probe. Thus the highest frequency that has sufficient penetration should be used for optimizing the image quality. Two to 7 MHz for transabdominal and 5 to 12 MHz for transvaginal probes are used for pelvic scanning.

US is a form of energy. It has two major effects in tissues it traverses: heating and mechanical bioeffects. The American Institute of Ultrasound in Medicine (AIUM) guidelines conclude that there is no independently confirmed evidence to indicate damage in animal models below a thermal index (TI) of less than 2 and mechanical index (MI) of less than 0.3. The as low as reasonably achievable (ALARA) principle should be followed for all US examinations, but in particular for first trimester examinations and use of Doppler techniques. 1

Normal Anatomy
The uterus, tubes, and the ovaries are located in the pelvis. The pelvic brim is bordered by the sacral promontorium and linea terminalis formed by the iliac arcuate line, iliopectineal line, and the pubic crest. The pelvic brim separates the bony pelvis into two compartments: the greater or false pelvis and the lesser or true pelvis. The true pelvis is the compartment caudal to the pelvic brim. The greater or so-called false pelvis is the upper part of the pelvis above the pelvic brim and is occupied by bowel. An enlarged bladder or pelvic masses may also extend to the greater pelvis. TAUS may be more helpful in those cases because TVUS is mostly limited to the true pelvis.

General Anatomic Descriptions

Uterus and the Fallopian Tube
The uterus is located between the bladder and the rectosigmoid colon. The uterus has two parts: corpus or body and cervix. The junction of the corpus with the cervix is called the isthmus . The fallopian tubes originate from the cornua of the uterus. The upper part of the corpus above the fallopian tubes is called the fundus . The anterior lower portion of the uterus is continuous with posterior wall of the bladder separated by a connective tissue layer. The upper anterior, lateral, and posterior uterine walls are covered by the peritoneum. The peritoneal folds between the bladder and the rectum are called anterior cul-de-sac and posterior cul-de-sac or pouch of Douglas . A mild amount of free fluid in the cul-de-sac is a variant of the normal ( Figure 1-1 ) ; however, increased amount of fluid in the posterior cul-de-sac and any fluid in the anterior cul-de-sac usually indicates pathology.

FIGURE 1-1 Mild amount of free fluid is seen on transvaginal imaging of the uterus, sagittal section (arrow) . The uterus is anteverted and anteflexed.
The uterine size and shape are related to age and parity. In term neonates, the length of the uterus is correlated with birth weight and ranges from 2.3 to 4.6 cm, with a mean of 3.4 cm. The cervix is larger than the fundus (fundus-to-cervix ratio = 1:2), and the maximum thickness is approximately 1.4 cm; the endometrial lining is often echogenic. A small amount of fluid in the cavity can be seen secondary to high estrogen levels before delivery. The prepubertal uterus has a tubular configuration; however, in some cases the anteroposterior dimension of the cervix is larger than the anteroposterior dimension of the fundus, with a spade shape. The endometrium can be visualized as a thin echogenic line using high-frequency transducers. The length of the prepubertal uterus is 2.5 to 4 cm, and the anteroposterior dimension of the uterus does not usually exceed 10 mm. 2
The uterus starts to grow before menarche and continues to grow for several years.
The pubertal uterus has the adult pear configuration with the fundus of equal size or larger than the cervix and measures 5 to 8 cm long, 3 cm wide, and 1.5 cm thick. The fundus-to-cervix ratio is 1:1 to 3:2 in nulliparous women and 3:2 to 2:1 in multiparous women. The dimensions of the uterus in nulliparous and multiparous women are 6 to 8.5 cm and 8 to 10 cm in length, 3 to 5 cm and 4 to 6 cm in width, and 2 to 4 cm and 3 to 5 cm in the anteroposterior dimension, respectively. The uterine size decreases after menopause with a tendency of the fundus-to-cervix ratio to decrease.
The cervix is fixed in the pelvis by ligaments, whereas the uterus can assume various positions. In the majority, the uterus is anteverted, that is, tilted anteriorly, at the cervicovaginal junction and sits on the bladder. Version describes the relationship of the cervix to the vagina. Flexion describes the relationship of the cervix to the uterus. A retroflexed uterus is tilted backward instead of forward at the junction of the cervix and body, that is, the isthmus. A retroverted uterus refers to a backward tilt of the entire uterus, including the cervix ( Figure 1-2 , A and B ).

FIGURE 1-2 A, Anteverted anteflexed uterus on transabdominal imaging in the sagittal plane. B, Retroverted retroflexed uterus in an 18-year-old patient on transvaginal imaging in the sagittal plane. Calipers show appropriate measurement of endometrial thickness. Mild amount of fluid (arrow) in the cul-de-sac. C, Transvaginal ultrasound in the sagittal plane demonstrates calcifications (arrows) in the arcuate vessels of the uterus of an 89-year-old patient.
The myometrium sometimes appears to have three zones, not necessarily well demarcated. The inner layer appears hypoechoic and thin. It is not always seen clearly. The middle layer appears homogenous and is the muscle layer arranged in spiral configuration. The outer layer is thin and separated from the middle layer by arcuate vessels, which are variably evident by US. Calcifications may be seen in the arcuate vessels of older women ( Figure 1-2 , C ).
The endometrium is composed of a basal layer and functional layer shed each month. The endometrial thickness is measured perpendicular to the longitudinal axis excluding the hypoechoic subendometrial area, that is, the inner myometrium, and should be measured at the thickest point (see Figure 1-2 , B ) including the thickness on each side of the endometrial cavity. If there is fluid in the endometrial cavity, the endometrium should be measured excluding the fluid rim ( Figure 1-3 ). The endometrium is poorly seen in the majority of the patients with a retroverted uterus on TAUS examination because of the angle and backward position of the fundus; TVUS can usually allow adequate endometrial evaluation in those cases. For a good quality endometrial thickness measurement, a well-defined distinct endometrial echo should be seen extending from the endocervical canal to the fundus (see Figure 1-2 , B ). When visualization is suboptimal as a result of fibroids, previous surgery, marked obesity, or an axial uterus, the endometrial echo should be reported as “poorly seen.” An axial uterus, also sometimes termed a midpositioned uterus, occurs when uterine long axis is in a partly retroverted position, such that the uterine long axis extends directly away from the probe. Saline infusion sonohysterography (SHG) and hysteroscopy are both appropriate next steps in the endometrial evaluation of such patients, if patients have a history of bleeding. The endometrial texture should be assessed, and if heterogeneous and irregular, this may be a more important determinant than endometrial thickness. Because endometrial carcinoma, polyps, and hyperplasia can be focal, the entire endometrium (from cornua to cornua) should be imaged in longitudinal and transverse views. 3

FIGURE 1-3 Transvaginal ultrasound in the sagittal plane demonstrates fluid (arrow) in the endometrial cavity of an 81-year-old patient. The endometrial measurements (calipers) appropriately exclude the endometrial fluid. The thickness of the two layers would then be summed for the endometrial thickness measurement.
The endometrium varies in appearance during the menstrual cycle ( Figure 1-4 ). The endometrium appears ultrasonographically as a thin, hyperechogenic single line immediately after menses in the early proliferative phase of the menstrual cycle. The slightly hyperechoic, well-defined endometrium gradually thickens up to approximately 8 mm. The functional and basal layers can be visually differentiated during the mid–late follicular phase. In the late follicular and periovulatory period, the endometrium assumes a trilaminar appearance with a central echogenic line of opposing functional layers, which are hypoechoic, and slightly hyperechoic basal layers more peripherally, and may measure up to 12 to 16 mm. A homogeneous, hyperechoic endometrium is observed as endometrial glands branch and expand under the influence of luteal progesterone production in the secretory phase. If pregnancy occurs, echogenicity and thickness are maintained as decidual reaction to implantation starts to progress. If pregnancy does not occur, the endometrium begins to regress in thickness, with echogenicity remaining similar or becoming heterogenous, finally ending in breakdown of the functional layer. An occasional tiny echogenic focus may sometimes be seen near the interface of the endometrium and myometrium. These are probably of no clinical significance and may be due to dystrophic calcifications from previous instrumentation.

FIGURE 1-4 A, Transvaginal ultrasound in the sagittal plane demonstrates the endometrium (calipers) measuring 5 mm in thickness on day 7, proliferative phase. B, Transvaginal ultrasound in the sagittal plane demonstrates trilaminar appearance of the proliferative phase endometrium on day 10. Endometrial thickness (calipers) measures 7.6 mm. C, Transvaginal ultrasound in the sagittal plane demonstrates uniform hyperechoic appearance of the late secretory phase endometrium, on day 31. The endometrial thickness (calipers) measures 12.9 mm.
In postmenopausal women with vaginal bleeding, large prospective studies have shown that an endometrial thickness of 4 mm or less on TVUS has a small risk for malignancy of 1 in 917 cases. Thus biopsy is not indicated in postmenopausal patients with bleeding when endometrial thickness is 4 mm or less. 3 In the postmenopausal patient without any bleeding, there is considerable disagreement for defining the upper limit of endometrial thickness, giving a range of 5 to 15 mm. 4, 5 The significance of a thick endometrial echo in nonbleeding postmenopausal women has not been validated in a prospective trial. Routine tissue sampling in this group is not recommended; however, a recent study in asymptomatic women with endometrial thickness of 6 mm or more reported a 3.1% risk for endometrial carcinoma using both hysteroscopy as well as dilation and curettage. 3, 6
Screening for endometrial pathology in women receiving hormone replacement is not recommended. The endometrial thickness shows a wide range of variation in asymptomatic women receiving estrogen alone or in some combination of estrogen and progestin with a mean endometrial thickness of 6.0 mm (range, 1 to 15 mm).
Tamoxifen-induced subepithelial stromal hypertrophy leads to poor correlation between endometrial pathology and ultrasonographic endometrial thickness with low specificity and with positive predictive values as low as 1.4%. 7
The cervix should be demonstrated on both transabdominal and transvaginal images although it is best seen on transvaginal images. The cervical palmate folds or plicae palmatae are the mucosal folds in the cervical canal and can sometimes be visualized by TVUS. Blockage of endocervical glands results in a frequent finding of nabothian cysts ( Figure 1-5 ). Nabothian cysts are usually of no clinical significance. Tiny echogenic foci may sometimes be seen centrally in the cervix. They have been found in association with chronic cervicitis and are also speculated to be due to dystrophic calcifications from previous instrumentation. They generally are of no clinical significance.

FIGURE 1-5 Nabothian cyst (arrow) in the cervix on transvaginal sagittal ultrasound image.
The fallopian tubes, or oviducts, extend outward from the superolateral portion of the uterus and end by curling around the ovary. The fallopian tubes connect the cornua of the uterine cavity and the peritoneal cavity. The oviducts are between 10 and 14 cm in length and slightly less than 1 cm in external diameter. The mesentery of the tubes, the mesosalpinx, contains the blood supply and nerves. Each tube is divided into four anatomic sections: The intramural or interstitial segment is 1 to 2 cm in length and is surrounded by myometrium. The isthmic segment begins as the tube exits the uterus and is approximately 4 cm in length. This segment is narrow, 1 to 2 mm in inside diameter, is straight, and has the most highly developed musculature. The ampullary segment is 4 to 6 cm in length and approximately 6 mm in inside diameter. It is wider and more tortuous in its course than other segments. Fertilization normally occurs in the ampullary portion of the tube. The infundibulum is the distal trumpet-shaped portion of the oviduct. Approximately 20 to 25 irregular finger-like projections, termed fimbriae, surround the abdominal ostia of the tube. One of the largest fimbriae is long enough to reach the ovary, the fimbria ovarica .
The interstitial portion of the fallopian tube may be seen on transverse uterine sections and on coronal 3D imaging. The other parts of the fallopian tubes are not usually visualized unless there is a pathologic process that enlarges the tube or in the presence of significant amount of fluid in the adnexal region ( Figure 1-6 ).

FIGURE 1-6 Transvaginal ultrasound shows a right hydrosalpinx, appearing as a typical elongated anechoic structure (arrows) with an incomplete septation.

Embryologic Remnants
Hydatids of Morgagni are common simple cysts that occur near the fimbriated ends of the fallopian tube. They are Wolffian duct remnants and can achieve the size of approximately 1 cm. They can rarely undergo torsion with infarction by strangulation of their mesentery. Paraovarian cysts may be Wolffian duct or paramesonephric duct remnants, with the latter occurring more commonly within the broad ligament rather than at the fimbriated ends of the fallopian tube. Paraovarian cysts are usually simple cysts that range from 1 to 8 cm, and are usually asymptomatic but can infrequently become symptomatic as a result of enlargement and/or torsion. Sonographically, one cannot generally distinguish hydatids of Morgagni from paraovarian cysts, and the distinction is usually clinically insignificant.

The Ovaries
The ovaries are a pair of oval-shaped glands, usually located lateral to the uterus and medial to the internal iliac vessels in the ovarian fossa. Their position can vary within the pelvis, however. The ovarian volume can be assessed by the formula for ellipsoids: height × width × depth × 0.52. The size of the ovaries varies depending on age and parity ( Figure 1-7 ). The ovarian volume among children up to 24 months old can be larger than 1 cm 3 , and small cysts or follicles may be observed. 8 The mean volume is approximately 1.1 cm 3 among girls up to 1 year of age and decreases to 0.67 cm 3 among girls 13 to 24 months old. Cysts larger than 9 mm can be seen in 18% of the ovaries in girls aged 1 day to 12 months. 8 The ovaries grow in size between the age of 2 and 14 years. The number of follicles larger than 5 mm increases from 7 to 9 years of age. During the reproductive years, ovaries measure approximately 1.5 × 2.5 × 4 cm. The average ovarian volume in premenopausal women is approximately 7.4 to 7.8 cm 3 (standard deviation [SD], 2.4 to 2.6) and is not influenced by parity. 5 The ovarian volume decreases in the postmenopausal patient showing a significant relationship with the number of years postmenopause. Average ovarian volume in the early postmenopausal patient is 3.4 to 3.8 cm 3 (SD, 1.3 to 1.6) and in late menopause is 2.5 cm 3 (SD, 1.1 to 1.3). 5

FIGURE 1-7 Normal adult ovary in longitudinal (long) and transverse (trv) plane on transvaginal ultrasound. The ovary is located medial to the internal iliac vessels. Calipers indicate ovarian measurements, and the ovarian volume is normal.
Human follicular development occurs from a diameter of approximately 0.03 mm and continues for more than 150 days until ovulation is achieved. Follicles are visible by US at only relatively advanced stages of development (i.e., ≥4 mm). They grow in minor and major waves of development, with smaller follicles appearing to grow and regress in a random fashion during the interovulatory interval. The growth dynamics of follicles up to 4 mm in diameter are not known. During the menstrual cycle, several ovarian follicles grow to 8 to 12 mm in diameter. The dominant follicle can be recognized, usually at a mean size of 10 mm, at day 8 to 12 of the menstrual cycle. It starts to differ from other follicles and increases in size 2 to 3 mm/day. The intermediate follicles usually grow to less than 15 mm. The growth of the dominant follicle continues and at the time of ovulation has a mean diameter ranging from 17 to 27 mm. In a menstruating woman, a simple ovarian cyst up to 25 mm most likely represents a follicle and should not be reported as a cyst. Disappearance or sudden decrease in size of the dominant follicle and appearance of free fluid in the cul-de-sac are the most sensitive markers of follicle rupture and ovulation. Irregularity of the follicle walls and internal echoes may also occur with follicle rupture, although they appear to have lower sensitivity and specificity for follicle rupture. 9 After release of the egg, the dominant follicle partially collapses, forming a corpus luteum. There may be some internal bleeding as a result of vascularization of the granulosa layer of the ovary after ovulation, forming a corpus hemorrhagicum. The corpus luteum atrophies to a corpus albicans that is not usually visible by US. When supporting a conceptus, the corpus luteum maintains its hormonal secretion during the first trimester. Its size remains static from 5 to 9 weeks’ gestation (mean, 17 mm) and gradually regresses with almost 20% undetectable by 10 to 13 weeks, when placental hormone production takes over.
Functional ovarian cysts are seen in reproductive women and fall into two categories: follicular and corpus luteum cysts. Follicular cysts arise when this physiologic release fails and follicular growth continues as a result of either excessive stimulation by follicle-stimulating hormone or from lack of the normal preovulatory luteinizing hormone surge. Follicular cysts rarely grow larger than 10 cm, and most are asymptomatic. Although central hemorrhage into the collapsed follicle after ovulation is normal, expansion of the cavity by hemorrhage is consistent with a hemorrhagic corpus luteum. At some size, the term hemorrhagic corpus luteum cyst may be appropriate, but that size is not clear. Hemorrhagic cysts can vary in size ranging from approximately 3 to 8.5 cm. The hemorrhagic cyst has enhanced through-transmission signifying the basic cystic nature of the mass, and has a wall of variable thickness. Fine reticular echoes representing fibrin strands with no blood flow on Doppler US are the most common appearance ( Figure 1-8 ). The cyst shows change in echo pattern with time, related to the temporal sequence of clot formation and lysis. Hemorrhagic ovarian cysts can be followed sonographically to spontaneous resolution in 6 to 8 weeks; most completely resolve in 6 weeks or decrease considerably in size and change in morphologic appearance. The various phases of the retraction of the blood clot can result in sonographic images that may mimic a fluid level or a papillary excrescence. Rupture of a hemorrhagic corpus luteum can occasionally happen and may mimic an ectopic pregnancy.

FIGURE 1-8 Transvaginal ultrasound demonstrates a right ovarian hemorrhagic cyst in a 21-year-old woman.

The vagina is a thin-walled, distensible, fibromuscular tube that extends from the vestibule of the vulva to the uterus. The bladder and urethra are anterior, and rectum is posterior to the vagina. The walls of the vagina are normally in apposition and flattened in the anteroposterior diameter forming an appearance of the letter H in cross-section. The axis of the upper portion of the vagina lies fairly close to the horizontal plane when a woman is standing, with the upper portion of the vagina curving toward the hollow of the sacrum. In most women an angle of at least 90 degrees is formed between the axis of the vagina and the axis of the uterus (see Figure 1-2 , A ). The lower third of the vagina is in close relationship with the urogenital and pelvic diaphragms. The middle third of the vagina is supported by the levator ani muscles and the lower portion of the cardinal ligaments. The upper third is supported by the upper portions of the cardinal ligaments and the parametria ( Figure 1-9 ).

FIGURE 1-9 Tissue support to the vagina is provided by urogenital and pelvic diaphragms in the lower third, by the levator ani muscles and the lower portion of the cardinal ligament in the middle third, and by the cardinal ligaments and the parametria in the upper third.

Pelvic Muscles
Pelvic muscles include lower limb muscles (psoas major, iliacus, piriformis, obturator internus), pelvic diaphragm (levator ani, coccygeus), and urogenital diaphragm.

Lower Limb Muscles
The iliacus muscle attaches to the inner side of the iliac pelvis. The psoas major muscle originates on the T12 through the L5, passes through the false pelvis, exits the pelvis, and inserts to the lesser trochanter merged with the iliacus muscle. The psoas and iliacus muscles can be visualized lateral to the external iliac vessels by US ( Figure 1-10 ). This combined muscle controls flexion of the hip.

FIGURE 1-10 Transabdominal ultrasound demonstrates the psoas (vertical arrow) and iliacus (horizontal arrow) muscles lateral to the external iliac vessels. The ovary is not visualized.
The piriformis muscle originates from the sacrum, the part of the spine in the gluteal region, and from the superior margin of the greater sciatic notch. It exits the pelvis through the greater sciatic foramen with sciatic nerve to insert on the greater trochanter of the femur. The obturator internus muscle originates on the medial surface of the obturator membrane, the ischium near the membrane, and the rim of the pubis. It exits the pelvic cavity through the lesser sciatic foramen. The obturator internus is situated partly within the lesser pelvis, and partly at the back of the hip joint. The piriformis and the obturator internus function as hip abductors and lateral hip rotators. The piriformis muscle can be visualized next to the sacral promontorium. The obturator internus muscle can be located at the lateral pelvic wall.

Pelvic Diaphragm
The pelvic diaphragm is a wide thin muscular layer of tissue that forms the inferior border of the abdominopelvic cavity. Composed of a broad, funnel-shaped sling of fascia and muscle, it extends from the symphysis pubis to the coccyx and from one lateral sidewall to the other. The levator ani and coccygeus muscles attached to the inner surface of the minor pelvis form the muscular floor of the pelvis ( Figure 1-11 ).

FIGURE 1-11 The levator ani and coccygeus muscles attached to the inner surface of the minor pelvis form the muscular floor of the pelvis.
The levator ani muscles are divided into three components named after their origin and insertion: pubococcygeus, puborectalis, and iliococcygeus. The inner border of the puborectalis muscle forms the margin of the levator (urogenital) hiatus, through which passes the urethra, vagina, and anorectum. The coccygeus is a triangular muscle that occupies the area between the ischial spine and the coccyx. When the body is in a standing position, the levator plate is horizontal and supports the rectum and upper two thirds of vagina above it. The opening within the levator ani muscle through which the urethra and vagina pass is called the urogenital hiatus of the levator ani . Levator ani weakness may loosen the sling behind the anorectum and cause the levator plate to sag. This opens the urogenital hiatus and predisposes to pelvic organ prolapse.

Urogenital Diaphragm (Perineal Membrane)
The urogenital diaphragm is a funnel-shaped sleeve of striated muscle external to the pelvic diaphragm and includes the triangular area between the ischial tuberosities and the symphysis ( Figure 1-12 ). The urogenital diaphragm covers the levator hiatus and is made up of the deep transverse perineal muscles, the constrictor of the urethra, and the internal and external fascial coverings. Anteriorly the urethra is suspended from the pubic bone by continuations of the fascial layers of the urogenital diaphragm. The free edge of the diaphragm is strengthened by the superficial transverse perineal muscle. Posteriorly the urogenital diaphragm inserts into the central point of the perineum. Situated farther posteriorly is the ischiorectal fossa. Located more superficially are the bulbocavernosus and ischiocavernosus muscles (see Figure 1-12 ). The urogenital diaphragm is important in urinary continence. Voluntary contraction closes the bladder neck and inhibits the bladder contraction.

FIGURE 1-12 The urogenital diaphragm is a funnel-shaped sleeve of striated muscle external to the pelvic diaphragm and includes the triangular area between the ischial tuberosities and the symphysis.

Perineal Body
The perineal body is a pyramidal fibromuscular structure situated in the midline between the vagina and the anus. The rectum, the pubococcygeus and perineal muscles, and the external anal sphincter attach to the perineal body. Acquired weakness of the perineal body gives rise to elongation and predisposes to defects such as rectocele and enterocele.

The broad ligaments are peritoneal folds that extend from the lateral margins of the uterus to the pelvic walls. The upper part of the broad ligament includes peritoneal folds that cover the oviduct, the uteroovarian ligament, and the round ligament. The fallopian tubes attach to the inner two thirds of the superior margin to form the mesosalpinx, to which the fallopian tubes are attached. The outer third of the superior margin forms the infundibulopelvic ligament or suspensory ligament of the ovary, through which the ovarian vessels traverse. The uterine vessels and the ureter are found in the inferior medial portion of the broad ligament .
The cardinal ligaments form the base of the broad ligaments, laterally attaching to the fascia over the pelvic diaphragm and medially merging with fibers of the endopelvic fascia. The cardinal ligaments are composed of connective tissue that medially is united firmly to the supravaginal portion of the cervix. The round ligaments extend from the lateral portion of the uterus and are located below and anterior to the origin of the oviducts. Each round ligament is covered by a fold of peritoneum that is continuous with the broad ligament and extends outward and downward to the inguinal canal, to terminate in the upper portion of the labium majus.
The uterosacral ligaments extend from their posterolateral attachment to the supravaginal portion of the cervix to encircle the rectum and insert into the fascia over the sacrum. The ligaments are composed of connective tissue and some smooth muscle and are covered by peritoneum. They form the lateral boundaries of the pouch of Douglas.

Pelvic Vessels
The common iliac artery passes laterally, anterior to the common iliac vein to the pelvic brim. The iliac artery is anteromedial to the iliac vein on the right and is anterolateral to the iliac vein on the left side. At the lower border of L5, the common iliac artery divides into internal and external iliac branches. The external iliac artery courses medial to the psoas muscle border in the false pelvis and gives off only two branches, the inferior epigastric artery and the deep circumflex iliac artery, and then, after passing under the inguinal ligament, becomes the femoral artery, which is the primary blood supply to the lower limb. The external iliac artery is lateral to the external iliac vein on the right and anteromedial to the external iliac vein on the left.
After passing over the pelvic brim, the internal iliac arteries divide into anterior and posterior trunks ( Figure 1-13 ). The posterior trunk consists of three branches: the iliolumbar artery, the lateral sacral artery, and the superior gluteal artery. These vessels are closely related to the nerve plexus on the piriformis muscle. The superior gluteal artery is the largest branch of the internal iliac artery and supplies the muscles and skin of the gluteal region.

FIGURE 1-13 Pelvic vessels.
The anterior trunk of the internal iliac artery has several branches, including the obliterated umbilical artery or the so-called medial umbilical fold, superior vesical, inferior vesical, uterine, middle rectal, obturator, internal pudendal, and inferior gluteal arteries. The uterine artery arises from the medial surface of the internal iliac and courses medially over the ureter in the base of the broad ligament to the uterus at the cervical level. The uterine artery then gives branches that anastomose with the ovarian and vaginal artery.
The vagina is supplied by a branch of the uterine artery on the anterior portion and by a branch from the internal iliac artery for its posterior portion. The ovaries receive their blood supply primarily from the aorta. The uterine artery gives off its ovarian and tubal branches in the mesosalpinx and the ovarian suspensory ligament at the upper edge of the broad ligament. These branches anastomose with the ovarian artery in the posterolateral border of the ovary. The uterine artery branches into arcuate arteries in the uterine wall. Arcuate arteries give rise to radial arteries that run parallel to the myometrium and supply the myometrium. Spiral arteries are branches of the radial arteries and supply the basal layer of the endometrium.
In ovulatory cycles, uterine artery blood flow varies cyclically, with perfusion being highest (and resistance to flow lowest) immediately before ovulation and in the midluteal phase. In general, blood flow increases in response to estrogens. There appears to be no difference in blood flow between the two uterine arteries in relation to which ovary contains the dominant follicle.

Bladder and Ureter
The bladder is located in the anterior lesser pelvis posterior to the symphysis pubis. The bladder is a hollow muscular organ that functions to store and evacuate urine. It is attached to the anterior abdominal wall by the urachus. The urachus is a solid cord of tissue that represents an obliterated embryologic canal. It connects the fetal bladder with the allantois, a structure that contributes to the formation of the umbilical cord. The bladder is covered by peritoneum on the superior aspect. Inferiorly, the bladder is attached to the pubic bone by dense condensations to the posterior aspect of the pubic bone, known as the pubovesical ligaments . On US examination, bladder walls should be uniform in thickness. Ureteral jets can be visualized and can be used to differentiate pelvic cystic masses from a full bladder.
The urethra can be seen on transperineal imaging and on transvaginal imaging with the probe partially inserted into the vagina until the bladder neck is visualized.

Indications for pelvic US:
1. Gynecologic indications, including pelvic pain, vaginal bleeding, palpable mass, and infertility evaluation
2. Obstetric indications, including pregnancy confirmation and dating, diagnosis and management of ectopic pregnancy, follicle monitoring
3. Suspected disease of other pelvic organs

Transabdominal (TAUS) and transvaginal (TVUS) approaches are the most common techniques for ultrasonographic pelvic assessment; however, transperineal or translabial and transrectal approaches, although less commonly, are used as well.
In general TAUS and TVUS techniques are used as complementary examinations. Patients who cannot tolerate TVUS can be candidates for TAUS only. TAUS can give better details for large pelvic masses with abdominal extension. Patients who cannot fill the bladder or who are having follicle monitoring can have a TVUS examination only.

Technique Description
TAUS examination is performed with the patient in the recumbent position and with a full bladder to expand the bladder to prevent bowel loops in the scanning area and to improve resolution as a result of fluid interface.
After the TAUS examination, patients should void to empty their bladder for TVUS. Information about the examination and verbal consent for internal examination should be obtained. The transvaginal probe is covered with two layers of condoms or other protective material as a precaution for rupture. US gel on the tip of the transducer and on the outside of the cover facilitate acoustic coupling with the transducer and vaginal wall. Air between the cover and the transducer should be avoided. Saline or water can be used as a lubricant if the US examination is being performed for infertility workup and in particular if a sperm insemination is planned after the examination to avoid the potential adverse effects on sperm motility with some types of lubricants.
Many patients prefer to insert the TVUS probe themselves, and their choice should be asked before probe insertion. A chaperone should be present if possible. TVUS is usually performed with the patient in the lithotomy position. The insertion should be watched as the probe is slowly advanced to the anterior fornix to identify the vagina, bladder, uterus, urethra, and rectum. Sagittal and transverse views of the uterus should be performed. The cervix and the cul-de-sac should be visualized. The probe is then angled at the level of the fundus toward the lateral pelvis on both sides near the iliac vessels to locate the ovaries, which should be imaged in sagittal and transverse planes. The transvaginal probe can be manipulated to assess pain and to clarify the origin and elasticity of any pelvic masses or to assess whether the pelvic organs are freely moving over the other tissue (sliding sign) to rule out adhesive disease. The free hand can be placed on the abdomen to simulate a bimanual examination using the probe instead of the second hand. The transvaginal probe is cleansed after use following the manufacturer’s requirements and rinsed before use to avoid chemical irritants.

Transperineal Technique
A midline sagittal view is obtained by placing a transducer (usually a 3.5- to 7-MHz curved array) on the perineum after covering the transducer with a glove or other protective material. Imaging can be performed in dorsal lithotomy position, with the hips flexed and slightly abducted, or in the standing position. A full or a half full bladder is preferred to visualize the bladder neck. The presence of a full rectum may impair diagnostic accuracy and sometimes necessitates a repeat assessment after defecation. Parting of the labia may be necessary to improve image quality. The transducer can usually be placed against the symphysis pubis without causing significant discomfort. The resulting image includes the symphysis anteriorly, the urethra and bladder neck, the vagina, cervix, rectum, and anal canal ( Figure 1-14 ). Transperineal scans are useful for assessment of obstructed uterovaginal anomalies such as primary amenorrhea with hematometra or hematometrocolpos. They may also help in the evaluation of Mayer-Rokitansky-Küster-Hauser syndrome and vaginal septa, as well as assessment of structural and functional integrity of the pelvic floor. 10

FIGURE 1-14 Transperineal rendered image of the urethra, vagina, and anal canal. Abdominal three-dimensional probe was placed on the perineum in the sagittal plane. Postprocessing performed using the original data from the sagittal plane, adjusting size of the rendering box to the region of interest.
TAUS and TVUS routine images and measurements generally include the following:
• Sagittal uterine views, with longitudinal and anteroposterior uterine measurements
• Transverse uterine views, with transverse uterine measurement at the fundus
• Cervical image in sagittal view revealing cul-de-sac and if necessary in transverse planes
• Measurement of endometrial thickness, excluding any fluid and subendometrial hypoechoic zones, at its thickest point, which is generally toward the fundus; measurement should be perpendicular to the longitudinal axis of the uterus and includes both layers of the endometrium, that is, the layer on each side of the endometrial cavity
• Bilateral ovarian views and measurements of the sagittal, anteroposterior diameters in parasagittal planes and transverse diameters
Any uterine masses or ovarian cysts or masses should be measured in three planes for size. Uterine size, myometrial echotexture, presence of fibroids or other uterine masses, the echotexture of the endometrium and thickness, ovarian appearance and presence of ovarian cystic or solid masses, any other pelvic masses, and presence of free fluid in the pelvis and/or abdomen should be documented.

Three-Dimensional Ultrasound
Three-dimensional US is a recent technologic improvement in medical sonography. The ability to rapidly acquire and store ultrasonographic data volumes is a potential major advantage in terms of examination time and retrospective analysis. Stored sonographic volume data can be manipulated to reveal several two-dimensional planes and 3D reconstruction of the particular volume data set. The technique enables the qualitative and quantitative assessment of sonographic volume data with the use of several analysis tools, such as multiplanar imaging, surface and volume rendering, and semiautomated and automated volume calculations. Three-dimensional US is still yet to be widely used on a routine basis. The proceedings of the AIUM Consensus Conference for use of 3D ultrasound in gynecology concluded that “3D ultrasound appears to be a problem-solving tool in selected circumstances and may well become a part of many obstetric and gynecologic ultrasound examinations in the future” and suggested the following as indications for a 3D examination in gynecology 11 :
1. Assessment for congenital anomalies of the uterus ( Figure 1-15 )
2. Intrauterine device location and type ( Figure 1-16 )
3. Mapping of myomata for planning myomectomy
4. Cornual ectopic pregnancies
5. Evaluation of the endometrium and uterine cavity with or without saline infusion SHG ( Figure 1-17 )
6. Imaging of adnexal lesions to distinguish ovarian from tubal origin and ovarian from uterine origin
7. Abscess drainage in the pelvis and abdomen
8. Three-dimensional guidance in interventional procedures for infertility
9. Evaluation and monitoring of patients with infertility, including patients with polycystic ovaries and tubal occlusion

FIGURE 1-15 Normal endometrium and fallopian tubes, uterine contour on reconstructed coronal image from three-dimensional ultrasound examination.

FIGURE 1-16 Reconstructed coronal image from three-dimensional ultrasound examination demonstrates Mirena intrauterine device (IUD) within the uterine cavity of a 26-year-old patient. IUD string also noted.

FIGURE 1-17 Three endometrial polyps on tomographic imaging of the coronal section on three-dimensional ultrasound during sonohysterography in a 33-year-old woman with heavy menometrorrhagia.

SHG is a procedure in which saline, sometimes warmed, is instilled into the uterine cavity to enhance endometrial visualization during TVUS. TVUS evaluation of the endometrium should precede SHG so that one knows the morphology of the endometrium before injecting saline. SHG can help diagnose submucosal leiomyomas and endometrial pathology, such as polyps, hyperplasia, cancer, and adhesions, and may help avoid invasive diagnostic procedures. 12 In premenopausal women with abnormal vaginal bleeding, SHG results will indicate the etiology as dysfunctional or secondary to a structural pathology and help tailor therapy. In postmenopausal women with abnormal vaginal bleeding, SHG can distinguish bleeding caused by atrophy (the most common cause of bleeding in this age group) from anatomic lesions that might require tissue sampling and/or resection for treatment (see Figures 1-17 and 1-18 ). SHG should be performed in the early follicular phase of the menstrual cycle, after cessation of menstrual flow and before day 10, because the endometrium is thin at this point in the cycle. The physiologically thicker endometrium in the secretory phase can give the false appearance of an endometrial polyp or hyperplasia. In infertility patients, there is also the possibility of infusing saline to an early pregnant uterus. Blood clots, intrauterine debris, mucus plugs, and shearing of normal endometrium by the catheter tip can also give false-positive results. SHG should not be performed in a woman with a pelvic infection or unexplained pelvic tenderness. If painful, dilated, or obstructed fallopian tubes are found before saline infusion, the examination should be delayed to administer antibiotics. In the presence of nontender hydrosalpinx, antibiotics may be given at the time of the examination. Active vaginal bleeding is not an absolute contraindication; however, the interpretation may be more challenging.

FIGURE 1-18 Normal sonohysterography sagittal section in a 46-year-old woman with menometrorrhagia.
Anesthesia or analgesia is not required for insertion of the intrauterine catheter because this is often painless. Sterile technique is preferable to prevent endometritis and other infections. The patient is placed in the lithotomy position, and a speculum is inserted into the vagina. The cervical os is localized and cleaned with a povidone–iodine solution or chlorhexidine gluconate. Intracervical or intrauterine catheters can be used. The catheter should be flushed with sterile saline before insertion to clear it of air, which can cause an echogenic artifact in the uterine cavity. The catheter is then inserted through the cervical os into the cervical canal. The speculum is then removed carefully to avoid dislodging the catheter. A dilator or guidewire can be used if there is difficulty passing the catheter through the cervical os. After the speculum is taken out, the vaginal probe is inserted. The whole uterine cavity is scanned systematically from one side to the other and from the bottom to the top. Available data are controversial whether 3D SHG is superior to 2D SHG in terms of agreement with hysteroscopy findings. 13, 14 Procedure-related adverse effects and complications are mild and uncommon. In a prospective study, the complication rates were reported as failure to complete the procedure, 7%; pelvic pain, 3.8%; vagal symptoms, 3.5%; nausea, 1%; and postprocedure fever, 0.8%. 15 Preprocedure administration of nonsteroidal antiinflammatory drugs does not appear to change the pain score compared with placebo.


The quality of a pelvic ultrasound examination is dictated by the correct selection of probes and the scanning experience of the sonologist or sonographer.
The highest frequency that has sufficient penetration enables optimal image quality. Settings of 2 to 7 MHz for transabdominal and 5 to 12 MHz for transvaginal probes are used for pelvic scanning.
Ultrasound has two major effects in tissues it traverses: heating and mechanical bioeffects, There is no independently confirmed evidence to indicate damage in animal models below a thermal index (T1) of < 2 and mechanical index (M1) of < 0.3.The ALARA (as low as reasonably achievable) principle should be followed for all ultrasound examinations.
Transabdominal and transvaginal routine images and measurements generally include the following:
• Sagittal uterine views, with longitudinal and anteroposterior uterine measurements.
• Transverse Uterine views, with transverse uterine measurements at the funds.
• Cervical image in sagittal view revealing cul-de-sac and if necessary in transverse planes.
• Measurement of endometrial thickness, excluding any fluid and subendometrial hypoechoic zones, at its thickest point.
• Bilateral ovarian views and measurements of the sagittal and anteroposterior diameters in parasagittal planes and tranverse diameters.
Any uterine masses or ovarian cysts or masses should be measured in three planes for size. Uterine size, myometrial echotexture, presence of fibroids or other uterine masses, echotexture of the endometrium and thickness, ovarian size, appearance and presence of ovarian cystic or solid masses, any other pelvic masses and presence of free fluid in the pelvis and/or abdomen should be documented.

Suggested Readings

1. Sliver P. Pelvic ultrasound in women. World J Surg . 2000;24:188-197.
2. Lindheim S.R., Morales A.J. Comparision of somohysterography and hysteroscopy: lessons learned and avoiding pitfalls. J Am Assoc Gynecol Laparosc . 2002;9:223-231.
3. Ghi T., Casadio P., Kuleva M., et al. Accuracy of three-dimensional ultrasound in diagnosis and classification of congenital uterine anomalies. Fertil Steril . 2009;92:808-813.
4. Santoro G.A., Wieczorek A.P., Dietz H.P., et al. State of the art: an integrated approach to pelvic floor ultrasonography. Ultrasound Obstet Gynecol . September 2, 2010. epub ahead of print


1. Ziskin M.C. Update on the safety of ultrasound in obstetrics. Semin Roentgenol . 1990;25:294-298.
2. Garel L., Dubois J., Grignon A., et al. US of the pediatric female pelvis: a clinical perspective. Radiographics . 2001;21:1393-1407.
3. Goldstein S. The role of transvaginal ultrasound or endometrial biopsy in the evaluation of the menopausal endometrium. Am J Obstet Gynecol . 2009;201:5-11.
4. Lin M.C., Gosink B.B., Wolf S.I., et al. Endometrial thickness after menopause: effect of hormone replacement. Radiology . 1991;180:427-432.
5. Merz E., Miric-Tesanic D., Bahlmann F., et al. Sonographic size of uterus and ovaries in pre- and postmenopausal women. Ultrasound Obstet Gynecol . 1996;7:38-42.
6. Schmidt T., Breidenbach M., Nawroth F., et al. Hysteroscopy for asymptomatic postmenopausal women with sonographically thickened endometrium. Maturitas . 2009;62:176-178.
7. Holbert T.R. Transvaginal ultrasonographic measurement of endometrial thickness in postmenopausal women receiving estrogen replacement therapy. Am J Obstet Gynecol . 1997;176:1334-1338.
8. Cohen H.L., Shapiro M.A., Mandel F.S., et al. Normal ovaries in neonates and infants: a sonographic study of 77 patients 1 day to 24 months old. AJR Am J Roentgenol . 1993;160:583-586.
9. Ecochard R., Marret H., Rabilloud M., et al. Sensitivity and specificity of ultrasound indices of ovulation in spontaneous cycles. Eur J Obstet Gynecol Reprod Biol . 2000;9:59-64.
10. Dietz H.P. Ultrasound imaging of the pelvic floor. Part I: two-dimensional aspects. Ultrasound Obstet Gynecol . 2004;23:80-92.
11. Benacerraf B.R., Benson C.B., Abuhamad A.Z., et al. Three- and 4-dimensional ultrasound in obstetrics and gynecology: proceedings of the American Institute of Ultrasound in Medicine Consensus Conference. J Ultrasound Med . 2005;24:1587-1597.
12. Elsayes K.M., Pandya A., Platt J.F., et al. Technique and diagnostic utility of saline infusion sonohysterography. J Gynaecol Obstet . 2009;105:5-9.
13. Opolskiene G., Sladkevicius P., Valentin L. Two- and three-dimensional saline contrast sonohysterography: interobserver agreement, agreement with hysteroscopy and diagnosis of endometrial malignancy. Ultrasound Obstet Gynecol . 2009;33:574-582.
14. Terry S., Banks E., Harris K., et al. Comparison of 3-dimensional with 2-dimensional saline infusion sonohysterograms for the evaluation of intrauterine abnormalities. J Clin Ultrasound . 2009;37:258-262.
15. Dessole S., Farina M., Rubattu G., et al. Side effects and complications of sonohysterosalpingography. Fertil Steril . 2003;80:620-624.
Chapter 2 Pitfalls in Gynecologic Ultrasound

Cheryl L. Kirby, Mindy M. Horrow
Ultrasound (US) imaging is prone to pitfalls related to technique, normal variations, and interpretative errors. Nowhere are these issues more problematic than in the female pelvis. In this anatomic region, numerous scanning approaches and transducers may be chosen. A range of findings may be normal or abnormal depending on the age of the patient, previous surgery, parity, medications, and the stage of the menstrual cycle. Lastly, one must consider pathologic processes with similar appearances and those in adjacent nongynecologic organs. In this chapter we will address a variety of recurrent pitfalls encountered over the years in our practice.

Scanning Technique and Related Pitfalls
Traditionally a patient presenting for pelvic sonography was requested to distend her urinary bladder. The distended bladder serves as a sonographic window for transabdominal imaging by enhancing the through-transmission of the sound beam and displacing gas-filled small bowel loops out of the pelvis to allow better visualization of the uterus and adnexae. Transvaginal imaging, which allows better resolution because of a higher frequency transducer, is best performed with an empty bladder. The competing requirements of transvaginal and transabdominal imaging, especially in patients without previous studies, often engender long waits, uncomfortable patients, frequent interruptions to empty a rapidly filling bladder, and a chaotic US schedule. As a result, several years ago we decided to forgo the requirement for a full bladder in most patients.
Our protocol begins with a brief transabdominal evaluation, surveying for a large uterus with exophytic fibroids, large adnexal masses or ovaries, and/or the uterus and ovaries displaced out of the pelvis. If none of these situations apply, we perform a transvaginal scan only. Occasionally transabdominal imaging alone is sufficient or even preferred, and sometimes a combination is required. The vast majority of our population can undergo a transvaginal examination. Children and teenagers are asked to distend their bladders and are scanned transabdominally.
Our preference for transvaginal scanning may result in pitfalls for the less experienced sonographer or sonologist. In some cases, transvaginal scanning only images the cervix and lower uterine segment, missing the majority of the uterine body and fundus ( Figure 2-1 ). This situation occurs frequently in patients with a previous cesarean section when the lower anterior uterus is tethered to the anterior abdominal wall at the site of the cesarean incision. This scarring elongates the cervix, pulling the uterine body out of the pelvis and beyond the range of the vaginal transducer. A similar problem is encountered when only transvaginal imaging is performed in a patient with a large myomatous uterus. In these situations transabdominal imaging is preferred. To help in planning the scan, we inquire about a history of cesarean section in all parous patients.

FIGURE 2-1 Limited area visualized on transvaginal scan. A, Sagittal diagram of the pelvis demonstrates the limited field of view (gray area) seen with transvaginal imaging in many patients with scarring from previous cesarean section. The scarring causes tethering of the lower uterine body to the anterior abdominal wall with resultant elongation of the cervix. This limits transvaginal visualization of the uterus to only the cervix and lower uterine segment thereby missing the majority of the uterine body and fundus, accounting for a recurrent pitfall. Additional transabdominal images are required in these patients. B, Sagittal transvaginal ultrasound image of a uterus in a patient with previous cesarean section shows limited field of view with visualization of the lower 8.7 cm of the uterine body and cervix and nonvisualization of the remainder of the uterine body and fundus. C, Sagittal transabdominal ultrasound in the same patient reveals full size of the uterus (12.9 cm in length) with tethering of the anterior lower uterine body. (The authors acknowledge the artistic contributions of Alyson Singer.)
We occasionally supplement or substitute with transperineal and transrectal imaging in women who cannot tolerate the lithotomy position or the vaginal transducer and in women who cannot maintain a full bladder. For transrectal imaging, the patient is placed in the left lateral decubitus position with flexed hips and knees. Generally this position and approach are well tolerated in the elderly population, allowing adequate visualization of the uterus and endometrium. Evaluation of the adnexal structures is often incomplete because the excursion of the transducer from side to side is more limited than on a routine transvaginal examination. Transperineal imaging is easily tolerated and helpful in evaluation of the vagina, cervix, and urethra.

Uterine Pitfalls

The Benign Enlarged Uterus
Fibroids occur in 20% to 30% of females more than 30 years old, accounting for many nongravid pelvic US referrals. Their diagnosis is usually straightforward. The most typical appearance is a hypoechoic solid mass with an internal whorled pattern, shadowing, and circumferential vessels. Rapid enlargement may lead to necrosis and degeneration and a variety of more atypical appearances with cystic or fatty components (lipoleiomyoma) and calcifications. Although most fibroids arise in the intramural portion of the uterus, less common locations may be subserosal, submucosal, cervical, and within the broad ligament. These variations in appearance and location may result in errors of diagnosis. A subserosal myoma with a thin attachment to the uterine body may be confused with a solid ovarian mass or even missed with transvaginal imaging. A myoma with significant cystic degeneration may simulate an ovarian cyst. 1 Pressure with the transvaginal probe and/or color Doppler imaging to demonstrate bridging vessels between the uterine body and pedunculated myoma may help in the correct diagnosis. A densely calcified myoma is usually not a diagnostic dilemma, but the shadowing may cause significant technical limitations in evaluation of the adnexa and the endometrium.
Usually multiple fibroids result in the classic US appearance of an enlarged, lobulated uterus with hypoechoic masses and variable amounts of posterior acoustic shadowing. Even when the posterior shadowing is dense, the anterior lobulated margin allows one to make the correct diagnosis. Somewhat confusing, however, is the patient with an enlarged uterus secondary to a large dominant fibroid. Frequently the large fibroid is mistaken for the entire uterus when this dominant fibroid displaces a smaller, more normal-appearing uterine corpus toward the periphery ( Figure 2-2 ). Distinguishing between a solitary enlarged myoma and multiple myomata may affect treatment options. One should be wary of this pitfall when presented with images of a uterus that appears as a single mass without any normal myometrium. A careful search for the endometrium and its connection to the endocervical canal using transabdominal and transvaginal techniques may help to avoid this error.

FIGURE 2-2 Dominant uterine fibroid mistaken as a diffusely myomatous main body of the uterus. A, Sagittal transabdominal ultrasound shows exophytic dominant fibroid mistakenly measured as the entire fundus and body of the uterus. B, Transverse ultrasound image reveals the displaced, relatively normal body of the uterus (arrows) with the exophytic fibroid (F) arising from the right side.
Diffuse or focal enlargement of the uterus may also be caused by adenomyosis. In adenomyosis the extension of endometrial glands into the myometrium causes smooth muscle hypertrophy accounting for enlargement of the uterus. Sonographic features include heterogeneity of the myometrium with cysts and hyperechoic foci, linear striations, thin non-edge shadows, penetrating vessels, and poor definition of the endometrial myometrial junction. The contour of the uterus is usually more globular and less lobulated than with multiple fibroids ( Figure 2-3 ).

FIGURE 2-3 Focal adenomyosis. A, Sagittal transvaginal ultrasound image reveals a smooth globular uterine contour with focal elliptical area of heterogeneity ( calipers ) in a retroverted uterus. B, Penetrating nondisplaced uterine vessels coursing into the region of adenomyosis. Color Doppler imaging helps to distinguish adenomyosis with penetrating vessels from a fibroid with typical circumferential vessels.
Distinguishing between focal fibroids and focal adenomyosis (adenomyoma) is more challenging and prone to potential errors ( Figure 2-4 ). Findings that favor an adenomyoma over a myoma include poorly defined margins, minimal mass effect on the adjacent endometrium, small cysts, ellipsoid versus spherical shape, presence of echogenic foci and striations, and the absence of calcifications, edge shadowing, and large vessels at the margin of the lesion. 2 With the improvement of US equipment in the past 10 to 15 years, we have seen several cases initially interpreted as fibroids that on subsequent studies are focal regions of adenomyosis. Because patients with fibroids and adenomyosis have similar symptoms, differentiation between the two diagnoses is important because treatment options differ. Finally, in cases with adenomyosis and fibroids, shadowing from the fibroids may limit evaluation of the presence and severity of the adenomyosis.

FIGURE 2-4 Focal adenomyosis in posterior uterine body myometrium confused for fibroid. A, On transabdominal ultrasound image, a focal hypoechoic region (arrows) was mistaken for fibroid. B, Transvaginal sagittal view of uterus confirms the area as focal adenomyosis in light of its poor definition (thick arrows) , myometrial cyst (thin arrow) , and pencil-thin edge artifact. Cursors mark endometrium. Small anterior uterine body fibroid also present.
An enlarged uterus may be palpated in patients with duplication anomalies such as a bicornuate or didelphys uterus because of the separation of the two uterine horns. When the endometrial canal is divided, the uterine fundus must be evaluated to differentiate a septate uterus from a bicornuate or didelphys uterus. Pregnancy outcomes and treatment options vary significantly for these different anomalies. Volumetric US imaging is helpful in differentiating these entities. Reconstruction of the true coronal plane of the uterus allows evaluation of the peripheral fundal contour and the extent of separation of the endometrial canals. A superior exophytic fibroid arising in an otherwise normal uterus may be confused for a bicornuate uterus because they have similar contours. The lack of a split endometrial canal and classic findings of a fibroid should help distinguish these two diagnoses.

False-Positive Diagnosis of Fibroids
Fibroids may calcify, but not all uterine myometrial calcifications are fibroids. Calcification of arcuate uterine arteries is common in elderly women, especially those with diabetes, and is considered part of the normal aging process. Arcuate artery calcifications usually appear as discontinuous linear parallel echoes ( Figure 2-5 ) located between the outer and intermediate layers of the uterine myometrium. 3 These calcifications must not be confused with calcification in uterine fibroids, which tend to be coarser and clumped and often associated with a noncalcified mass. Intense shadows from arcuate vessel calcification may obscure the endometrium and make it difficult to distinguish from the leading edge of a calcified myoma. Magnetic resonance may be the only imaging method for evaluation of the endometrium in this situation.

FIGURE 2-5 Calcification of the uterine arcuate arteries. Sagittal transvaginal ultrasound of the uterus reveals calcifications in a discontinuous parallel line configuration (arrows) located in the mid to outer aspect of the uterine myometrium.
A frequent site for an erroneous fibroid diagnosis is the lower uterine segment in a patient with a history of previous cesarean section. The cesarean section incision is usually transversely oriented in the lower uterine segment where there is increased fibrous tissue available for healing, thereby reducing the chance of future dehiscence. Not infrequently, distortion and prominence of the overhanging tissue superior to the scar ( Figure 2-6 ) assumes a rounded, relatively hypoechoic appearance, mistaken for a fibroid ( Figure 2-7 ). Thurmond et al. 4 suggest that differences in muscle contraction may account for the thicker superior edge of the scar that becomes more pronounced with increasing numbers of cesarean sections. Histologic evaluation reveals congestion of the endometrium above the scar, which may also account for this prominent tissue. 5

FIGURE 2-6 Distortion of lower uterine segment from previous cesarean section. Sagittal image from transabdominal ultrasound reveals typical, prominent anterior lower uterine segment, tethered to anterior abdominal wall (arrow) .

FIGURE 2-7 Two different patients with cesarean section scars mistaken for fibroids. A, Sagittal transvaginal ultrasound image reveals rounded configuration of the tissue superior to the cesarean section scar (S) . The arrow points to the bulbous tissue (mistaken for a fibroid) located superior to the scar. B, Sagittal transvaginal ultrasound image reveals edge artifact (thick arrow) originating from scar and rounded configuration of the tissue adjacent to the scar (thin arrow) .
The presence of posterior edge shadowing adds to the similarity of this scar tissue to a focal fibroid. This artifact manifests as narrow, vertical hypoechoic lines originating along the lateral margin of a rounded structure. 6 Edge artifact is common along the border of a cesarean section scar because of the rounded configuration of the myometrium adjacent to the scar. The hypoechoic edge artifact may be mistaken for posterior acoustic shadowing, further confusing this configuration for a fibroid ( Figure 2-7 , B ).

The endometrial thickness and appearance must be correlated with the menstrual status to assess for benign or malignant causes of bleeding such as polyps, fibroids, hyperplasia, and cancer. Accurate identification of the endometrium may be challenging in patients with multiple fibroids. Anterior fibroids that significantly attenuate the sound beam may completely obscure the endometrium. In addition, one must avoid mistaking tissue heterogeneity within fibroids or tissue interfaces between fibroids as the endometrium ( Figure 2-8 ).

FIGURE 2-8 Tissue interface between myometrial fibroids mistaken as the endometrium. A, Cursors mark the erroneous measurement of the echogenic tissue interface between two fibroids as the endometrium on this transabdominal sagittal ultrasound image of the uterus. B, Cursors correctly measure the endometrium that can be followed into the lower uterine segment.
The endometrium is measured in the midline sagittal plane of the uterus. If the uterus is oriented obliquely or coronally, a true sagittal plane may be impossible to obtain with routine imaging. Oblique images of the endometrium may falsely increase its measurement and result in overdiagnosis of endometrial thickening. Frequently this problem is worse on transvaginal imaging and may be overcome on transabdominal images, which allow for more scanning planes. Three-dimensional volume acquisition can correct for uterine obliquity by allowing visualization of the endometrium in a multitude of reconstructed planes regardless of the original acquisition plane. 7
Faulty measurement of the endometrium may occur in patients with adenomyosis, especially with transvaginal imaging. The increased resolution of the transvaginal transducer allows for improved visualization of the ectopic endometrial tissue located in the myometrium. The striations or patches of ectopic tissue cause poor definition of the endometrial myometrial junction ( Figure 2-9 ) and may cause pseudowidening of the endometrium ( Figure 2-10 ).

FIGURE 2-9 Adenomyosis causing indistinct endometrial borders. Focal area of adenomyosis blurs the margin of the endometrium (area between arrows ) on this sagittal transvaginal ultrasound image. Note the well-defined endometrium in areas without adenomyosis.

FIGURE 2-10 Pseudothickening of the endometrium in patient with adenomyosis. A, Sagittal transvaginal ultrasound image reveals erroneous thickened measurement of the endometrium (measured as 3.5 cm) by including the endometrial tissue and the adjacent echogenic changes in the subendometrial myometrium from adenomyosis. B, Sagittal transabdominal ultrasound image of the uterus reveals true measurement of the endometrium (3 mm) and adjacent heterogeneity (arrows) in the anterior uterine myometrium representing adenomyosis.
A potentially missed cause of dysfunctional uterine bleeding occurs in menstruating women with a previous cesarean section. Such patients may experience several days of postmenstrual spotting with old blood. 4 This delayed bleeding results from either retained menstrual blood or in situ bleeding in an endometrial cavity pouch. 5 This niche is created from postoperative scarring that puckers the endometrium anteriorly and creates a small reservoir for the blood. The slow leakage of this blood may result from poor contraction of the uterine muscle around the scar. Hysterosonography may be helpful in identifying a uterine niche. Although not all such patients are symptomatic, awareness of this entity can be useful in patients with the appropriate symptoms ( Figure 2-11 ).

FIGURE 2-11 Fluid within cesarean section scar niche. Transvaginal sagittal ultrasound image of the cervix reveals a small amount of fluid retained in the endometrial niche (arrows) created from retraction of the tissues at the cesarean section scar.

Although the cervix receives great attention during sonography of the gravid uterus, it tends to be overlooked in the nongravid uterus. The cervix is located deep in the pelvis and thus transabdominally is positioned far from the US transducer. Visualization of the cervix is further limited by the lack of an acoustic window in women with nondistended bladders.
Cervical pathology can even be missed on a transvaginal examination. This pitfall may be due to the normal cervical heterogeneity, numerous nabothian cysts, or sonologist complacency, knowing that the cervix is examined annually for cancer with direct visualization and cytologic assessment. Detailed evaluation of the cervix, however, should be part of the pelvic US examination because abnormalities of the cervix, including polyps, fibroids, and inflammation, can account for vaginal bleeding. In particular, polyps protruding through the cervical os are frequently overlooked possibly because of their close proximity to the transducer. The lower cervix can be better evaluated by pulling the transducer away from the exocervix or using a transperineal approach.

Pitfalls in the Adnexal Regions
Basic US imaging of the adnexal regions often consists only of an assessment of each ovary for a dominant cyst or mass. Unfortunately, this approach may result in overlooking abnormalities that are extraovarian and extrauterine in origin. In addition, the ovary must be evaluated in the context of the patient’s age and menstrual history to determine whether the size and number of follicles are appropriate. One should not necessarily consider an ovary to be normal based on the lack of a cyst or mass. High-frequency transvaginal imaging with color, power, and/or spectral Doppler is essential for these evaluations. This section shall address a variety of diagnoses that may be overlooked or underdiagnosed in the adnexal regions.

Tubular Structures in the Pelvis
Many tubular structures are normally found in the pelvis, including fallopian tubes, veins, arteries, and bowel, including the appendix. Abnormalities of any of these “tubes” can cause symptoms for which pelvic US is performed. Thus it is important to determine the type of tubular structure and whether it is abnormal. Unfortunately, many sonologists and sonographers who notice an odd “tubular structure” will ascribe it to bowel and thus beyond their consideration.
It is usually assumed that the normal fallopian tube cannot be imaged routinely. Actually it is relatively easy to find portions of the normal fallopian tubes, especially if there is some free pelvic fluid and the examiner relaxes the pressure of the transvaginal probe ( Figure 2-12 , A ). The segments of the fallopian tube increase in thickness from uterus to fimbria and consist of the intramural portion, isthmus, and ampulla. Thus it is not uncommon to visualize the ampullary portion of the tube in cross section as a round or ovoid echogenic structure, 5 mm or less, adjacent to the ovary (see Figure 2-12 , B ). Often one appreciates the finger-like projections of the fimbria at the end of the fallopian tube. In a normal, nondilated tube, the lumen should not be visible.

FIGURE 2-12 Normal fallopian tube in two different patients. A, Transvaginal ultrasound image of an elongated view of a normal fallopian tube (arrows) with a Morgagni cyst (C) , which most frequently occurs at the fimbriated end of the tube. This fallopian tube is easily visualized because of adjacent free fluid. B, Transverse transvaginal ultrasound image reveals oblique view of the normal fallopian tube (arrows) between the uterus (U) and ovary (O) .
It is not uncommon to visualize a small simple cyst, not arising within or connected to the ovary. Usually this is a paratubal cyst, also known as a hydatid of Morgagni. These cysts arise from remnants of the müllerian duct located below the fallopian tube, usually near the fimbria (see Figure 2-12 , A ). Clinically they are insignificant and rarely symptomatic unless they undergo torsion and infarction.
When a fallopian tube dilates or becomes inflamed, it should be more easily identified. Fallopian tube dilatation usually implies obstruction, although the converse is not necessarily true. Although pelvic inflammatory disease (PID) is the most common cause of dilatation, other etiologies include endometriosis and adhesions from an inflammatory process such as ruptured appendicitis. Dilatation can be acute or chronic with classical sonographic findings based on the chronicity of disease.
The hallmark of PID is the abnormal fallopian tube. Usually caused by a sexually transmitted infection, the bacteria ascend from the cervix, through the uterus, and into the fallopian tubes. Initially with salpingitis, the wall of the tube thickens and the endosalpingeal folds may be visualized. 8 When the lumen occludes, the tube will dilate, filling with complex fluid that may be uniform, heterogeneous, or in levels ( Figure 2-13 ). In the acute stage the wall is thick, 5 mm or more, and when viewed in cross-section may demonstrate the “cogwheel sign.” 9

FIGURE 2-13 Pyosalpinx. Transvaginal ultrasound image reveals a thick-walled dilated fallopian tube containing a fluid/debris level of purulent material related to pelvic inflammatory disease. Note the “cogwheel sign” of the thickened endosalpingeal folds (arrow) .
In the chronic phase of PID, a hydrosalpinx may develop. The fluid within the tube becomes anechoic; the wall measures less than 5 mm; and a cross-sectional view may show “beads on a string” resulting from the short, thick endosalpingeal folds projecting into the lumen. In both acute and chronic cases of tubal dilatation, a useful marker of a cystic structure as the fallopian tube is the “incomplete septum” sign. This appearance is created when the dilated tube falls back on itself, and two walls are adjacent resulting in a linear echogenic protrusion arising from one side, but not reaching the opposite one, and thus not a true septation ( Figure 2-14 ). A simple hydrosalpinx may be misinterpreted as a cystic, septated ovarian tumor. If the ovary cannot be distinguished separately, it is important to use signs such as the incomplete septum sign or waist sign as markers for the fallopian tube. 10 We have also found that unlike a tumor or ovarian cyst, a chronic hydrosalpinx is often easily compressible with the vaginal transducer. Occasionally a peritoneal inclusion cyst may have an appearance similar to a hydrosalpinx. The inclusion cyst, which represents peritoneal fluid trapped by adhesions around an ovary, will usually be distinguished by thin septations, but no true wall and the lack of an incomplete septum sign.

FIGURE 2-14 Hydrosalpinx with “incomplete septum sign.” Sagittal transvaginal ultrasound image reveals a dilated simple fluid-filled tubular structure with an incomplete septum (arrow) characteristic of a dilated fallopian tube.
The other common cause of a dilated fallopian tube is a tubal ectopic pregnancy. The most common location of an ectopic pregnancy is the fallopian tube. Sonographic appearances include a gestational sac with or without a yolk sac and/or an embryo, a “donut sign,” or a more amorphous echogenic “mass.” If there is bleeding into the fallopian tube, the acute hematosalpinx may appear as a large, somewhat amorphous mass or collection separate from the uterus and ovary. When the ectopic pregnancy is early or there is only a small amount of bleeding, the tubular shape is more easily appreciated.
Dilated veins, or pelvic varices, are a common cause of tubular structures in the adnexal region. These varices should not be ignored because they may lead to pelvic congestion syndrome, a common but frequently underdiagnosed cause of pelvic pain ( Figure 2-15 ). The typical patient is multiparous with worsening of dull pelvic pain after standing or activity. The pelvic varices may be due to incompetent valves from multiple pregnancies but may also be secondary to portal hypertension or an obstructed inferior vena cava. The varices can be confirmed with color Doppler US, although occasionally due to slow flow, even spectral Doppler US will be negative. In these cases the slow-moving echoes will be visible on gray scale, helping to differentiate these veins from dilated fallopian tubes. The varices often connect to dilated arcuate veins in the periphery of the uterus causing uterine enlargement and tenderness. If symptomatic, the patient can be treated by percutaneous coil embolization or laparoscopic ligation of the varices.

FIGURE 2-15 Pelvic varices. A, Transvaginal ultrasound image reveals a dilated tubular structure with fine internal echoes, and a fluid/debris level, representing slow-moving blood. Initially structure was misinterpreted as a pyosalpinx. B, Imaging with sensitive color Doppler settings confirms vascular nature.
The other major source of tubular structures in the pelvis is bowel. Bowel is identified by observing the typical “gut signature” of alternating echogenic and hypoechoic layers corresponding to the layers from mucosa through serosa and resulting in a “bull’s eye” appearance. Peristalsis can be used to confirm small bowel. Simple dilatation of small and/or large bowel can be due to a primary process (obstruction, inflammation, ischemia) or a reactive ileus. US is useful to differentiate normal and abnormal peristalsis and may be the first clue to a bowel obstruction.
More focal abnormalities of bowel can suggest a specific diagnosis. Thus a blind-ending bowel loop larger than 6 mm may be identified in acute appendicitis. Crohn’s disease ( Figure 2-16 ), diverticulitis, lymphoma, and intussusception may all be suggested using US. Because it is a real-time interactive examination, US can help determine the site of pain or a rigid, aperistaltic segment of bowel.

FIGURE 2-16 Crohn’s disease. Transvaginal ultrasound reveals a dilated thick-walled tubular structure with gut signature (Bowel) representing the inflamed terminal ileum adjacent to normal right fallopian tube (FT) and right ovary with small cyst (O) .

Abnormal Ovaries Without a Mass or Dominant Cyst
Familiarity with the normal appearance of the ovaries from infancy through menopause is essential if one hopes to make diagnoses related to inappropriate ovarian size or number of follicles. The ovaries are often larger at birth with an average volume of 1 cm 3 and may be palpable as a result of multiple cysts, related to the influence of maternal hormones ( Figure 2-17 ). Such an appearance is not worrisome and will resolve during infancy. In infancy the ovarian volume is usually less than 1 cm 3 , increasing gradually with age to 2 to 3 cm 3 approaching menarche, continuing to increase in size during the teens with an average volume of 8 cm 3 . 11 Tiny ovarian follicles are common throughout childhood. In our experience, normal ovaries are largest during the 20s and 30s, often beginning to decrease in volume and number of visible follicles during the 40s and particularly as the woman approaches menopause. The mean ovarian volume in adult women is reported as 9.8 ± 5.8 cm 3 . After menopause the ovaries atrophy and follicles disappear. Size is related to time since menopause with volumes varying between 1 and 5 cm 3 . Ovarian volume more than 10 cm 3 in a postmenopausal woman is considered abnormal by some authors. 12

FIGURE 2-17 Normal newborn ovary. Transabdominal ultrasound image reveals multiple small cysts in an enlarged ovary secondary to stimulation from maternal hormones.
Evaluation of ovarian size and appearance is extremely important in patients with primary or secondary amenorrhea. Abnormally small ovaries with few if any follicles occur with chromosomal anomalies such as Turner’s syndrome, hypogonadism, and premature ovarian failure. Usually there is no visible ovarian tissue in patients with classical XO Turner’s syndrome. With XO mosaicism, however, small ovaries are more common. 13 Young nulliparous women on long-term oral contraceptives may also have relatively small ovaries.
Premature ovarian failure may be a cause of primary or secondary amenorrhea. It is typically associated with elevated follicle-stimulating hormone (FSH) levels. Approximately half of all cases are idiopathic. Most other causes are immunologic, with a small percentage as a result of chromosomal abnormalities. 14 US can be helpful in this diagnosis, with two thirds of patients demonstrating small ovaries with volumes similar to postmenopausal ovaries and few, if any, follicles.
Ovarian enlargement, either unilateral or bilateral without a focal lesion, has a significant differential diagnosis. The most common normal cause of unilateral enlargement is a corpus luteum, with or without a cystic component. This diagnosis should be straightforward by correlating the phase of the menstrual cycle with the typical low resistance circumferential flow in color and spectral Doppler US.
Unilateral enlargement with acute pain is usually due to ovarian torsion. Although there is often an underlying mass such as cystic teratoma, cystadenoma, or hemorrhagic cyst, torsion may occur in an otherwise normal ovary. In the early stages of torsion, the ovary will enlarge significantly. The stroma may appear heterogeneous as a result of hemorrhage and infarction. Increased number and size of follicles is typical, usually displaced to the periphery of the ovary. Often a careful search of the adnexal region will reveal a twisted ovarian pedicle resulting in the “whirlpool” sign.
Complete lack of arterial and venous flow in the affected ovary is the hallmark of ovarian torsion. A variety of technical pitfalls, however, may result in lack of flow, including inadequate depth of penetration, improper Doppler gain, and inappropriately elevated pulse repetition frequency. Conversely, flow is often detected with torsion and is helpful to predict viability of the ovary. Because the ovary derives a dual blood supply from the ovarian and uterine arteries, arterial flow may still be present, although usually asymmetrically decreased compared with the contralateral ovary. Lack of venous flow in the symptomatic ovary is a more useful diagnostic indicator than the absence of arterial flow. 7
A related but much less common entity is massive ovarian edema. Most patients are young women who present with acute pain, usually on the right side. On sonography the involved ovary is significantly enlarged and relatively solid appearing with multiple small peripheral follicles and normal Doppler flow ( Figure 2-18 ). Some of these cases are caused by subtotal torsion, and if blood flow is present, the ovary may be successfully untwisted at surgery. Some cases seem to be caused by a hemorrhagic corpus luteum. Usually patients are treated symptomatically. 15

FIGURE 2-18 Massive ovarian edema in a 46-year-old woman with right lower quadrant pain. A, Transvaginal ultrasound image reveals an enlarged right ovary (volume of 28.2 cc), with several tiny peripheral follicles. B, Prominent low resistance arterial flow was typical of a corpus luteum. At laparoscopy, there was no torsion.
Bilateral ovarian enlargement is most commonly due to polycystic ovarian syndrome. This complex syndrome is due to hyperandrogenism resulting in an elevated luteinizing hormone (LH)/FSH ratio. The hormonal imbalance causes chronic anovulation and infertility and a wide variety of other endocrinologic associations, including obesity, insulin resistance, and hirsutism. Because of the wide spectrum of clinical appearances and to exclude other causes of elevated androgens, pelvic US is frequently performed. Current international consensus standards for polycystic ovaries include 12 or more total follicles measuring 2 to 9 mm in diameter or increased ovarian volume (>10 cm 3 ). Although increased stroma and increased stromal echogenicity are common findings, they are considered more subjective and thus not required for diagnosis. The typical appearance, however, may only occur in half of the patients, and up to one third may have normal ovarian volumes. Usually, but not always, the ovarian enlargement is bilateral. 16
Other less common associations of bilateral ovarian enlargement include PID, pelvic varices, and ovarian hyperstimulation syndrome. In PID, oophoritis may result in enlargement without other abnormalities. Sometimes the ovarian size is overmeasured by inclusion of a slightly enlarged, nondilated fallopian tube in the measurement. Pelvic varices have been reported in association with enlarged, cystic ovaries possibly related to venous stasis. Ovarian enlargement is a frequent complication of ovulation induction and may result in ovarian hyperstimulation syndrome. The size of the ovaries and number and size of the cysts increase with the severity of the syndrome and may be complicated by ascites, pleural effusions, and hypovolemia.

Pitfalls in Imaging of Ovarian Lesions
Although many ovarian abnormalities have a typical sonographic appearance, there are several lesions that are more likely to be misdiagnosed or overlooked during US imaging. Occasionally a large, simple adnexal cyst can simulate the bladder, especially if the patient has just voided completely. This problem is more likely with transabdominal imaging ( Figure 2-19 ). In general, however, the most problematic lesions are solid ovarian masses, including cystic teratomas (dermoids) and stromal tumors.

FIGURE 2-19 Ovarian cyst mistaken as bladder. Transabdominal sagittal ultrasound image reveals an anterior simple ovarian cyst (C) mistaken for the bladder. Note the decompressed bladder (B) located inferior to the cyst.
The echogenic “plug” or Rokitansky protuberance is the most characteristic feature of a cystic teratoma. The plug, which consists of variable amounts of fat, calcification, hair, and soft tissue, can be variable in size. When the teratoma consists entirely of a plug, it will be completely echogenic with some posterior shadowing ( Figure 2-20 ). This type of teratoma, even when large, is the most easily missed because it blends in with air-filled bowel and mesenteric fat ( Figure 2-21 ). Dermoids with a cystic component are much more easily appreciated. Occasionally large dermoids may extend well above the uterus and be missed on transvaginal imaging. Displacement of a dermoid out of the pelvis may also occur with ovarian torsion.

FIGURE 2-20 Missed ovarian dermoid. A, The lateral echogenic fat-containing dermoid (arrows) was missed on this transvaginal pelvic ultrasound. Only the normal medial ovarian tissue (O) was recognized. B, Computed tomography with intravenous contrast confirms the fat-containing right ovarian dermoid (arrow) .

FIGURE 2-21 Large missed ovarian dermoid. A, On sagittal transabdominal ultrasound image, the region of increased echogenicity with shadowing (arrows) superior to the uterus was interpreted as bowel. This area was not well visualized on transvaginal images (not shown). B, Corresponding sagittal reconstructed computed tomographic image with intravenous contrast reveals a large dermoid (arrows) superior and anterior to the uterus.
Dermoids can also be overdiagnosed. The appearance of a dermoid may be simulated by a dilated appendix containing an appendicolith, acute hemorrhage into a cyst, and occasionally a lipoleiomyoma (an uncommon fat-containing myoma) ( Figure 2-22 ). 17 Lipoleiomyomas can usually be identified by noting their location within the uterus.

FIGURE 2-22 Lipoleiomyoma. A, Transabdominal ultrasound of the uterus reveals an intramural round echogenic mass with posterior acoustic shadowing. B, The intrauterine mass (arrows) has high signal intensity on the axial transverse T1-weighted magnetic resonance image and low signal intensity on the sagittal T1-weighted fat saturation sequence (not shown), confirming a fat-containing intrauterine mass.
Small echogenic ovarian foci are usually unrelated to germ cell tumors. Numerous tiny echogenic foci smaller than 5 mm are common, representing psammomatous calcifications, hemosiderin, or specular reflections from tiny cysts below the spatial resolution of the transducer ( Figure 2-23 ). Slightly larger focal calcifications, usually 1 cm or smaller, are also common. Retrospective studies have not shown interval development of neoplasms. 18

FIGURE 2-23 Echogenic foci in a postmenopausal ovary. Transvaginal ultrasound image reveals tiny benign echogenic foci (arrows) related to specular reflections from tiny cysts, calcifications, or hemosiderin.
Stromal–sex cord tumors of the ovary are less common than epithelial neoplasms and germ cell tumors. Because they are often solid, they may not be appreciated as ovarian in origin. This is particularly true of the fibroma–thecoma tumors ( Figure 2-24 ). These tumors are composed of fibrous cells with varying amounts of thecal cells and thus are often similar in appearance to uterine myomata. We have found two helpful maneuvers to differentiate between uterine and ovarian origin. Intermittent pressure on the mass with the vaginal transducer may accentuate the origin of a mass. Color and pulsed Doppler can confirm uterine origin by showing blood vessels extending from the uterus into an exophytic myoma ( Figure 2-25 ).

FIGURE 2-24 Large solid stromal ovarian tumor mistaken as myomatous uterus. A, Sagittal transabdominal ultrasound shows large ovoid hypoechoic solid mass initially interpreted as an enlarged myomatous uterus (cursors). Endometrium was not seen. B, Downward angled, sagittal transabdominal image reveals inferiorly displaced uterus with normal endometrium (arrow) and superior mass (M) C, On sagittal T2-weighted magnetic resonance imaging, the solid mass (M) is separate from the inferiorly displaced normal uterus (UT) .

FIGURE 2-25 Exophytic uterine fibroid with bridging vessels from the uterus. On this transvaginal ultrasound image, the solid mass (arrow) located to the right of the uterus (UT) could potentially be mistaken as adnexal. The bridging vessels originating from the uterus confirm the uterine origin.

US is typically the initial and often the only imaging study necessary for innumerable clinical symptoms and diagnoses in the female pelvis. Its advantages of portability, relatively low cost, real-time interaction with the patient, and lack of ionizing radiation result in a ubiquitous pattern of use by many practitioners. Modern day equipment allows for exquisite imaging using transvaginal, transabdominal, transperineal, and transrectal approaches with supplemental color, power, and spectral Doppler US. US imaging, however, is subject to many technical and interpretive pitfalls. The sonologist must combine a thorough knowledge of normal pelvic imaging throughout the menstrual cycle and the life cycle, a complete appreciation of pelvic pathology and experience with real-time scanning to derive the greatest benefit from pelvic sonography. Nonetheless, mistakes and pitfalls will occur. This article describes many of our experiences and advice for avoiding problems. Ultimately, another imaging study may be required. In general we favor magnetic resonance for problem solving because of its multiplanar capabilities, variety of imaging sequences without or with contrast, and lack of ionizing radiation.


Brief initial transabdominal imaging of the pelvis is required because transvaginal scanning alone may miss the superior uterine body and fundus in patients with previous cesarean section scarring or large myomatous uteri (see Scanning Technique and Related Pitfalls section).
Findings that favor focal adenomyosis over a myoma include poorly defined margins, minimal mass effect on the adjacent endometrium, small cysts, elliptical versus globular shape, echogenic foci and striations, absence of calcifications, and presence of penetrating, rather than peripheral, vessels (see Benign Enlarged Uterus section).
False-positive diagnosis of a fibroid frequently occurs in the region of cesarean section scarring from the prominent superior overhanging tissue and resultant edge artifact (see section False-Positive Diagnosis of Fibroids).
Pseudothickening of the endometrium may occur on transvaginal imaging because of an oblique orientation of the uterus or in patients with adenomyosis (see Endometrium section).
A dilated fallopian tube is identified by the incomplete septum sign on longitudinal images and short thick endosalpingeal folds projecting into the lumen on axial sections (see Tubular Structures in the Pelvis section).
Knowledge of the normal ovarian appearance from infancy through menopause is essential to diagnose inappropriate ovarian size or number of follicles (see Abnormal Ovaries Without a Mass or Dominant Cyst section).
Solid ovarian masses such as dermoids and stromal tumors are more frequently overlooked than cystic ovarian masses (see Pitfalls in Imaging of Ovarian Lesions section).


1. Baltarowich O.H., Kurtz A.B., Pennel R.G., et al. Pitfalls in the sonographic diagnosis of uterine fibroids. AJR Am J Roentgenol . 1988;151:725-728.
2. Reinhold C., Tafazoli F., Mehio A., et al. Uterine adenomyosis: endovaginal US and MR imaging features with histopathologic correlation. Radiographics . 1999;19:S147-S160.
3. Atri M., de Stempel J., Senterman M.K., et al. Diffuse peripheral uterine calcification (manifestations of Monckeberg’s arteriosclerosis) detected by ultrasonography. J Clin Ultrasound . 1992;20:211-216.
4. Thurmond A.S., Harvey W.J., Smith S. Cesarean section scar as a cause of abnormal vaginal bleeding: diagnosis by sonohysterography. J Ultrasound Med . 1999;18:13-16.
5. Morris H. Surgical pathology of the lower uterine segment cesarean section scar: is the scar a source of clinical symptoms? Int J Gynecol Pathol . 1995;14:16-20.
6. Steel R., Poepping T.L., Thompson R.S., et al. Origins of the edge shadowing artefact in medical ultrasound imaging. Ultrasound Med Biol . 2004;30(9):1153-1162.
7. Andreotti R.F., Fleischer A.C., Mason L.E. Three-dimensional sonography of the endometrium and adjacent myometrium. J Ultrasound Med . 2006;25:1313-1319.
8. Horrow M.M., Rodgers S.K., Naqvi S. Ultrasound of pelvic inflammatory disease. Ultrasound Clin . 2007;2(2):297-309.
9. Timor-Tritsch I.E., Lerner J.P., Monteagudo A., et al. Transvaginal sonographic markers of tubal inflammatory disease. Ultrasound Obstet Gynecol . 1998;12(1):56-66.
10. Patel M.D., Acord D.L., Young S.W. Likelihood ratio of sonographic findings in discriminating hydrosalpinx from other adnexal masses. AJR Am J Roentgenol . 2006;186:1033-1038.
11. Sonographic imaging of the paediatric female pelvis. Eur Radiol . 2005;15:1296-1309.
12. Pavlik E.J., DePriest P.D., Gallion H.H., et al. Ovarian volume related to age. Gynecol Oncol . 2000;77(3):410-412.
13. Haber H.P., Ranke M.B. Pelvic ultrasonography in Turner syndrome: standards for uterine and ovarian volume. J Ultrasound Med . 1999;18(4):271-276.
14. Falsetti L., Scalchi S., Villani M.T., et al. Premature ovarian failure. Gynecol Endocrinol . 1999;13(3):189-195.
15. Umesaki N., Tanaka T., Miyama M., et al. Sonographic characteristics of massive ovarian edema. Ultrasound Obstet Gynecol . 2000;16:479-481.
16. Balen A.H., Laven J.S., Tan S.L., et al. Ultrasound assessment of the polycystic ovary: international consensus definitions. Hum Reprod Update . 2003;9:505-514.
17. Hertzberg B.S., Kliewer M.A. Sonography of benign cystic teratoma of the ovary: pitfalls in diagnosis. AJR Am J Roentgenol . 1996;167:1127-1133.
18. Brown D.L., Laing F.C., Welch W.R. Large calcifications in ovaries otherwise normal on ultrasound. Ultrasound Obstet Gynecol . 2007;29:438-442.

Suggested Readings

Andreotti R.F., Shadinger L.L., Fleisher A.C. The sonographic diagnosis of ovarian torsion: pearls and pitfalls. Ultrasound Clin . 2007;2:155-166.
Horrow M.M., Rodgers S.K., Naqvi S. Ultrasound of pelvic inflammatory disease. Ultrasound Clin . 2007;2(2):297-309.
Reinhold C., Tafazoli F., Mehio A., et al. Uterine adenomyosis: endovaginal US and MR imaging features with histopathologic correlation. Radiographics . 1999;19:S147-S160.
Part Two
Computed Tomography
Chapter 3 Computed Tomography
Normal Anatomy, Imaging Techniques, and Pitfalls

Lejla Aganovic

Ultrasound and magnetic resonance imaging (MRI) are the preferred modalities when evaluating patients with suspected gynecologic pathology. Computed tomography (CT), however, continues to have an important role in imaging female patients and is commonly done when evaluating patients with abdominal and pelvic pain. CT also provides important information in patients with gynecologic malignancies, for both initial staging and further management of the disease. This chapter describes the CT anatomy of the normal female pelvis, which is necessary to ensure accurate evaluation of pelvic abnormalities.

Description of Technical Requirements
During the past 10 years, CT technology has developed significantly with the introduction of multidetector CT (MDCT) scanners. Compared with single-slice scanners, MDCT scanners have multiple detectors in the scanning direction, allowing for acquisition of more than one image per x-ray tube rotation. The major improvements of MDCT technology include the ability to reconstruct an image at various thicknesses different from the thickness set during image acquisition, decrease in scanning time, and wider scan coverage. MDCT can acquire isotropic scan data that allow reformatting images in any plane with spatial resolution identical to the original scanning plane.

Normal Anatomy

Size, shape, and position of the uterus depend on age, hormonal status, pregnancy, and the degree of the bladder distention. In women of reproductive age, the uterus is 6 to 9 cm long ( Figure 3-1 ). After menopause the size of the uterus significantly decreases ( Figure 3-2 ). When the uterus is anteverted ( Figure 3-3 ), it is visualized posterior and superior to the bladder, whereas the retroverted uterus projects into the cul-de-sac. The uterus is divided into the body and cervix. The cervix is typically located in the midline with the uterine body often deviating to one side of the pelvis. Although it is often difficult to clearly distinguish the uterus from the cervix, the two can be separated from one another by their configurations because the uterus is somewhat triangular in shape, whereas the cervix has a more rounded appearance. The cervix enhances to a lesser degree compared with the uterus and often appears hypodense (see Figure 3-1 ).

FIGURE 3-1 Normal uterus. Contrast-enhanced computed tomography (CT) scan through the pelvis of a 27-year-old woman shows normal uterus (U) with hypoenhancing endometrium (arrows) and hypoenhancing cervix (C) . Because of the anteverted position of the uterus, endometrial canal is visualized on axial CT image in its entire length. Right ovary contains corpus luteal cyst (arrowhead) .

FIGURE 3-2 Postmenopausal uterus. Computed tomography scan of a 69-year-old woman shows atrophic uterus (arrow) that enhances very faintly.

FIGURE 3-3 Anteverted uterus. Sagittal reconstruction computed tomographic image of a 34-year-old woman shows anteverted uterus (U), vesicouteral pouch (arrow), and rectovaginal pouch (arrowhead) .
The uterus is covered by peritoneum that anteriorly becomes separate from it at the level of the cervix, forming the vesicouterine pouch that lies between the uterus and bladder (see Figure 3-3 ). Posteriorly the peritoneum covers the surface of the uterus to the level of the dorsal vaginal fornix and then ascends along the anterior surface of the rectum, creating the rectouterine pouch (pouch of Douglas or cul-de-sac) (see Figure 3-3 ). 1
The endometrium is often seen as a central hypodensity, most commonly ovoid or triangular in shape, better delineated on contrast-enhanced images. The lower attenuation of the endometrium relative to the myometrium is normal for premenopausal patients and should not be mistaken for fluid ( Figure 3-4 ). This appearance is likely related to a less rich vascular supply of the endometrium relative to the myometrium. Although there are no currently established CT criteria to assess the endometrium, it has been shown that CT is relatively insensitive for detecting mild endometrial thickening but is better at detecting gross thickening. 2 Additionally, review of CT often leads to a false-positive conclusion for true endometrial thickening; therefore transvaginal sonography should always be used to confirm the CT findings. 2 Occasionally, because of anteversion or retroversion, the uterus is imaged along its true coronal plane on axial CT images. In these cases, the normal endometrium has a triangular shape on axial CT images ( Figure 3-5 , A ) and can be misinterpreted as thickened. When such a finding is seen, sagittal reconstruction images should be interpreted in addition to the axial images because they can typically confirm the absence of true endometrial thickening (see Figure 3-5 , B ). 2

FIGURE 3-4 Normal endometrium. Contrast-enhanced computed tomographic image through the uterus of a 24-year-old woman shows normally enhancing uterus. Endometrium is seen as central area of low attenuation (arrow) compared with densely enhancing myometrium (arrowheads) .

FIGURE 3-5 Contrast-enhanced computed tomography (CT) of normal endometrium. A, Axial CT image demonstrates prominent triangular-shaped endometrium (arrows) B, Sagittal reconstruction CT image of the same patient shows normal thin endometrium (arrows) .
On postcontrast images the uterus can have different enhancement patterns that depend on individual variables, most importantly the patient’s age and menopausal status. Enhancement patterns are transitory and most commonly occur on images obtained 30 to 120 seconds after contrast administration, with the uterus becoming homogeneous on more delayed imaging. Zonal patterns of enhancement that have been described include subendometrial ( Figure 3-6 ), peripheral myometrial ( Figure 3-7 ), and faint diffuse myometrial enhancement (see Figure 3-2 ). 3 Knowledge of these normal findings is helpful when unusual uterine enhancement is seen during routine studies.

FIGURE 3-6 Normal pattern of uterine enhancement. Computed tomographic image through the uterus at 50 seconds after intravenous administration of contrast shows avid enhancement of subendometrial area (arrows) .

FIGURE 3-7 Normal pattern of uterine enhancement. Computed tomographic image through the uterus at 80 seconds after the start of contrast injection shows predominant enhancement of the subserosal area (arrows) .
The main blood supply to the uterus is through the uterine artery , a branch of the internal iliac artery. The uterine artery reaches the uterus through the cardinal ligament and then divides into an ascending and a descending branch ( Figure 3-8 ). The uterine artery gives off multiple branches to the uterus as it courses between the layers of broad ligament, before anastomosing with the uterine branch of the ovarian artery. The uterine artery also gives off branches to the cervix, vagina, fallopian tubes, and ovary.

FIGURE 3-8 Pelvic vascular supply.
Congenital uterine anomalies, also known as müllerian duct anomalies, result from nondevelopment or partial or complete nonfusion of müllerian ducts resulting in various abnormalities involving the uterus, cervix, and vagina. The prevalence of these congenital anomalies is estimated to be up to 1%. MRI and three-dimensional (3D) ultrasound are currently imaging modalities of choice because of their high sensitivity in detection and high specificity in characterization of müllerian duct anomalies. Occasionally the diagnosis can be made on CT images, usually seen as incidental finding ( Figure 3-9 ). CT can detect the abnormal external contour of the uterus, which is present in the bicornuate ( Figure 3-10 ) or didelphys uterus.

FIGURE 3-9 Septate uterus. Computed tomographic image shows normal convex external contour of the uterine fundus (arrow) and a septum (S) . The intercornual angle between the distal ends of the horns is less than 60 degrees.

FIGURE 3-10 Bicornuate uterus. Computed tomographic image shows two symmetric and widely splayed uterine horns (arrows) with a deep fundal cleft.

Understanding the normal CT appearance of the cervix and parametria is essential when assessing cervical pathology. Although no strict criteria for the size of cervix have been established, the cervix is generally 2 cm long and less than 3 cm in diameter. The size of the cervix, however, is variable, depending on many factors, including hormonal status and pregnancy. Younger patients can have a normal cervix that is larger than 4 cm in diameter ( Figure 3-11 ). Similar to the uterus, the enhancement pattern of cervix can be zonal, although the cervix typically enhances more slowly compared with the uterus, and this finding should not be misinterpreted as abnormal (see Figure 3-1 ). The cervix consists of two parts, the supravaginal and pars vaginalis, a lower portion that protrudes into vaginal canal. When a tampon is present high in the vagina, it can be displaced to one side within the pars vaginalis, and this normal appearance should not be mistaken for a cervical or vaginal mass ( Figure 3-12 ). Small amounts of gas can occasionally be seen in the endocervical canal, likely secondary to voiding. Inclusion cysts of the cervix (nabothian cysts) can also be seen in the cervix, particularly if they are larger than 1 cm.

FIGURE 3-11 Normal cervix. Computed tomographic image of a 24-year-old female shows normal cervix (C), which measures 4.7 cm in diameter. Multiple vessels in the paracervical plexus are visualized (arrows) .

FIGURE 3-12 Tampon in vagina. Computed tomographic image of a 38-year-old woman shows tampon (arrow) in the vagina that is displaced to the right side by normal pars vaginalis of the cervix (not shown). Also note enhancing vaginal mucosa (arrow) .
The lateral margins of the cervix are outlined by the parametria. The parametria ( Figure 3-13 ) are diamond-shaped regions just lateral to the cervix. The parametria are composed of connective tissue that contains fatty elements and are located between the leaves of the broad ligament. Medially the parametria are contiguous with the uterus, cervix, and proximal vagina, and inferiorly with the cardinal ligament. Laterally they extend to the pelvic sidewalls. The parametria normally contain thin strands of soft tissue that likely represent small blood vessels, lymphatics, and fibrous tissue. These should be differentiated from stranding caused by malignant extension, which are typically thicker than 3 to 4 mm in diameter.

FIGURE 3-13 Normal parametria. Computed tomographic image through the cervix shows normal parametria (asterisk) lateral to the cervix (C) . Strands of soft tissue course through the parametria, representing blood vessels and lymphatics (arrows) .

The ovaries are ovoid parenchymal structures that most commonly contain soft tissue stroma with small cystic areas that represent normal follicles. Their appearance varies with age and hormonal status. In women of childbearing age, the average ovarian volume is 9.8 cm 3 ; in postmenopausal women, 5.8 cm 3 ; and in the premenarchal group, 3.0 cm 3 . 4 In menstruating women, the normal ovary can be identified on CT in most instances ( Figure 3-14 ). Postmenopausal ovaries may be difficult to detect on CT because of their small size and lack of cysts ( Figure 3-15 ). Ovaries are intraperitoneal structures and are always located within the peritoneum. In nulliparous women, they are typically located within the ovarian fossa, a shallow peritoneal depression, which is bounded anterolaterally by the external iliac arteries and posteriorly by the pelvic ureter (see Figure 3-14 ). 1 The position of the ovaries is variable due to laxity of the ovarian and suspensory ligaments, which anchor the ovary. Normal ovaries may be seen in the cul-de-sac, pelvic inlet, iliac fossa, and lower abdomen ( Figure 3-16 ). During the first pregnancy, the ovaries are displaced superiorly in the abdomen because of the enlarging uterus. After the delivery, the ovaries often do not return to their original position.

FIGURE 3-14 Normal ovaries. Contrast-enhanced computed tomographic image of a 27-year-old woman shows normal ovaries in ovarian fossa (arrows) , posterior to the external iliac vessels (arrowheads) and lateral to the uterus (U) . Both ovaries contain multiple follicles.

FIGURE 3-15 Postmenopausal ovaries. Computed tomographic image of a 78-year-old woman shows small ovaries (arrows) in ovarian fossa.

FIGURE 3-16 Unusual position of the left ovary. Computed tomographic image of a 22-year-old woman shows relatively high position of the left ovary (arrow) on psoas muscle (P), close to anterior abdominal wall.
Although the ovaries often contain small cysts, larger benign ovarian cysts are also commonly seen in premenopausal women, which in most instances represent physiologic changes. In premenopausal women a common finding is a corpus luteal cyst. These cysts are normal physiologic ovarian structures that form in the dominant follicle after ovulation. Corpus luteal cysts are typically smaller than 3 cm and have a thick, crenulated, and hyperenhancing wall (see Figures 3-1 and 3-17 ). 5 It is important to be familiar with a normal CT appearance of the corpus luteal cysts to prevent inaccurate diagnosis and unnecessary further workup. If a cyst appears complex on CT, further characterization by ultrasound should be performed to better assess internal complexity and to exclude a cystic ovarian neoplasm.

FIGURE 3-17 Corpus luteal cyst. Computed tomographic image of a 36-year-old woman shows normal uterus and ovaries (arrowheads) . Right ovary contains a cyst that has a thick, crenulated, and enhancing wall, consistent with corpus luteal cyst (arrow) .
The ovarian blood vessels travel through suspensory ligament to reach the mesovarium and enter the ovary. The mesovarium is a short peritoneal fold that attaches the ovary to the broad ligament. The ovarian artery arises from the aorta just below the renal arteries. The left ovarian vein drains into the renal vein, and the right ovarian vein drains directly into the inferior vena cava below the level of the renal veins. Tracking the ovarian vein from the level of the renal vessels inferiorly can be helpful in localizing the ovaries and especially in differentiating between ovarian and nonovarian masses ( Figure 3-18 ). The presence of this finding has been described as the ovarian vascular pedicle sign. 6

FIGURE 3-18 Computed tomographic scan of a patient with bilateral ovarian teratomas. Ovarian blood vessels are seen posterior to the tumor (arrows) . The vessels merge with the tumor through suspensory ligament of ovary (arrowhead), confirming the ovarian origin of the mass.

Ovarian Vein Reflux
Reflux of contrast material from the left renal vein into the left ovarian vein is commonly seen in asymptomatic patients and can be present in up to 50% of multiparous women. 7 Depending on the grade of reflux, retrograde opacification can reach the parauterine veins, and some patients can develop parauterine varices ( Figure 3-19 ). Although the reflux is most commonly seen in asymptomatic patients, it can be the sign of the pelvic congestion syndrome if the patient has chronic pelvic pain for at least 6 months without any identifiable organic cause.

FIGURE 3-19 Ovarian vein reflux. Computed tomographic image of a 39-year-old multiparous woman shows reflux of contrast material into the dilated parauterine veins (arrows) .

The vagina surrounds the inferior portion of the cervix creating anterior, posterior, and lateral fornices. These can be identified in the axial and sagittal planes. A small amount of gas can occasionally be seen in the vagina and should not be interpreted as abnormal. If the vagina is distended with large amount of gas or fluid, obstruction and fistula should be considered. On postcontrast imaging the vaginal mucosa enhances intensely ( Figure 3-20 ) in contrast to the vaginal wall, which enhances to a lesser degree.

FIGURE 3-20 Normal vagina. Contrast-enhanced axial computed tomographic image through lower vagina demonstrates densely enhancing vaginal mucosa (arrow) .

Several fascial condensations referred to as ligaments support the female genital organs and can be seen on CT examinations ( Figure 3-21 ). 1

FIGURE 3-21 Pelvic ligaments and viscera.
The broad ligament ( Figure 3-22 ) is formed by two layers of peritoneum that drape over the uterus and extend laterally to the pelvic sidewalls. At the superior free edge, the two layers enclose the fallopian tube. The extreme lateral part of the tube (the ampulla and the fimbria) is not enclosed. The lower margin of the broad ligament ends at the cardinal ligament. Between the two layers of the broad ligament is parametrium, which contains the fallopian tube, round ligament, ovarian ligament, uterine vessels, ovarian vessels, nerves, lymphatics, and portion of the ureter. The broad ligament is rarely seen on CT unless ascites is present; however, its position can be determined by the structures it contains or abuts.

FIGURE 3-22 Broad ligament. Computed tomographic axial image through pelvis shows large amount of ascites. Broad ligament (arrows) is seen as a thin strand of tissue extending laterally from the uterus to the pelvic sidewalls.
The round ligament ( Figure 3-23 ) consists primarily of fibromuscular tissue that attaches to the anterolateral uterus. It is located between the layers of the broad ligament, in front of and below the fallopian tube. This ligament passes through the inguinal canal and terminates in the labia majora. It is frequently seen on CT as a thin soft tissue density that extends laterally from the uterine fundus and gradually tapers as it courses toward the inguinal canal.

FIGURE 3-23 Round ligament. Computed tomographic image through the uterus shows normal round ligaments (arrows) that extend laterally from the uterine fundus toward the inguinal canal.
The cardinal ligament (transverse cervical ligament) ( Figure 3-24 ) forms the base of the broad ligament and provides the primary ligamentous support for the uterus and upper vagina. It represents the border between the parametrium and paravaginal tissues. The cardinal ligament passes laterally from the cervix and upper vagina to merge with the fascia overlying the obturator internus muscle. There is a wide variation in shape, contour, and thickness of this ligament; however, it commonly appears triangular with tapered or squared-off ends. When the uterus is deviated toward the one side of the pelvis, this ligament becomes more apparent. Asymmetric cardinal ligaments may be present as a normal anatomic variant, and in some women, this ligament is not seen as a discrete structure but instead as an irregular network of blood vessels, nerves, and connective tissue. Understanding the normal spectrum of this ligament is important when assessing for the presence of parametrial invasion in cervical cancer.

FIGURE 3-24 Cardinal ligament. Computed tomographic image obtained at the level of cervix shows asymmetric appearance of the cardinal ligaments, a common finding. Right cardinal ligament has triangular appearance (arrow) and tapers gradually toward the pelvic sidewall. Left one is not well seen.
The uterosacral ( Figure 3-25 ) ligaments are commonly seen on CT. They represent folds of the peritoneum that extend from the lateral cervix and course posteriorly toward the anterior body of the sacrum at S2 or S3. Its fibers fuse medially with the posterior fibers of the cardinal ligament and form a midline raphe.

FIGURE 3-25 Uterosacral ligament. Computed tomographic image through the cervix shows uterosacral ligament (arrow) extending from the posterolateral cervix toward the sacrum.
The ovarian ligament ( Figure 3-26 ) and suspensory ligament of the ovary (see Figure 3-18 ) are rarely seen on CT. The ovarian ligament extends medially from the ovary to the uterus. The suspensory ligament of the ovary occupies the lateral aspect of the free upper edge of the broad ligament and contains the ovarian artery and vein.

FIGURE 3-26 Ovarian ligament. Computed tomographic image through the uterus (U) and left ovary (O) shows ovarian ligament (arrow) that lies between the ovary and the uterus.

Lymph Nodes
CT is the most commonly used imaging modality in the workup of oncology patients. Evaluation of the lymph nodes has important therapeutic and prognostic significance in these patients. Pelvic lymph nodes can be divided into nodes that are located in the retroperitoneum adjacent to the pelvic sidewall and those that are adjacent to the pelvic viscera. The lymph nodes along the pelvic sidewall are divided into the common, external, and internal iliac (hypogastric) group and parallel the named arteries and veins ( Figure 3-27 ). Obturator lymph nodes medial to the foramen belong to the external iliac group and are important because they can be the first site of disease extension ( Figure 3-28 ). The normal pattern of lymphatic drainage is presented in Table 3-1 .

FIGURE 3-27 Diagram shows pelvic and paraaortic lymph nodes.

FIGURE 3-28 Pelvic lymph nodes. Computed tomographic image of a patient with cervical cancer shows enlarged left obturator lymph node (arrow) . Normal bilateral external iliac nodes with fatty hila are also seen (arrowheads) .
TABLE 3-1 Normal Lymph Node Drainage of Pelvic Organs Inguinal nodes Vulva, distal vagina, distal rectum, anus Internal iliac nodes Almost all pelvic organs External iliac nodes Bladder, proximal vagina, uterus, ovary, cervix Common iliac nodes Rectum, drainage from internal and external iliac nodes Obturator nodes Bladder, cervix Paraaortic and caval nodes Ovary, fallopian tube
The CT criteria for classifying the lymph nodes as harboring tumor are based on the size and morphology. Most benign lymph nodes have a central fatty hilum. The upper limits of normal for size are 7 mm for the internal iliac, 8 mm for the obturator, 9 mm for the common iliac, and 10 mm for the external iliac lymph nodes, measured in short axis diameter. 8 Coronal and sagittal reformations can be helpful for more accurate measurement of the shortest axis. Using size as the main criteria for identifying metastatic lymph nodes has its limitations. Lower thresholds are more sensitive to metastatic disease but result in more false-positive results, whereas higher thresholds improve specificity but have more false-negative results. This obstacle has been somewhat overcome by adding positron emission tomography (PET) imaging to CT, which increases specificity and sensitivity for detection of malignant lymph nodes.

Posthysterectomy Patient
CT is commonly performed in women who have undergone previous hysterectomy. A CT examination can be done in the immediate postoperative period to exclude hemorrhage or infection. When there is a high clinical suspicion for intraoperative injury to the ureter, an intravenous urogram or CT urogram can be done to evaluate for the presence of contrast extravasation from the ureters.
In women who have a remote history of hysterectomy, CT shows a vaginal cuff that often appears as symmetric soft tissue posterior to the bladder ( Figure 3-29 ). Presence of ovaries varies with the surgical procedure and is often unknown to the patient and her referring physician. Remaining ovarian tissue may be more difficult to identify in the absence of the uterus because of small size; however, it is important to identify normal ovaries, usually by tracing the gonadal vein or round ligament, to avoid mistaking them for lymph nodes or other masses. Ovarian vein thrombosis is a common incidental finding and can occur in up to 80% of cases in postoperative patients if hysterectomy and oophorectomy have been performed for malignant disease ( Figure 3-30 ). 9 These patients have not been reported to be at risk for embolic complications, unlike ovarian vein thrombosis associated with pregnancy.

FIGURE 3-29 Normal vaginal cuff. Computed tomographic scan through the lower pelvis shows normal vaginal cuff as bilateral, symmetric region of soft tissue (arrows) .

FIGURE 3-30 Ovarian vein thrombosis in a patient with history of ovarian cancer, hysterectomy, and bilateral salpingo-oophorectomy with lymph node dissection. Coronal reconstruction computed tomographic image of the abdomen shows the entire length of the filling defect in the right ovarian vein (arrows), which is often seen as an incidental finding in these patients.

Surgical Transposition of the Ovaries
Transposition of the ovaries is a surgical procedure performed in premenopausal patients who must undergo pelvic irradiation but wish to preserve their fertility. The ovaries are mobilized on a vascular pedicle and transposed outside the radiation field, most commonly lateral to the paracolic gutter or lateral to the colon in the iliac fossa ( Figure 3-31 , A ). One or two surgical clips are placed on the ovaries to help identify their location. The transposed ovary appears on CT as a soft tissue mass, commonly containing small physiologic cysts. Identification of the ovarian vessels leading toward the transposed ovaries aids in diagnosis (see Figure 3-31 , B ). Lack of familiarity with the CT appearance of transposed ovaries may lead to misdiagnosing the ovary as a metastatic deposit or peritoneal cystic lesion. A transposed ovary may also show increased fludeoxyglucose uptake on PET imaging (see Figure 3-31 , C ) because of functional changes, and the presence of this finding should not be misinterpreted as a metastatic deposit.

FIGURE 3-31 Surgically transposed ovary in a young patient with cervical cancer. A, Computed tomographic (CT) image through the midabdomen shows the right ovary that is transposed lateral to the ascending colon. The ovary contains a corpus luteal cyst (arrowhead) B, Coronal reconstruction CT image shows a metallic clip (arrow) that is surgically placed as a marker for the transposed ovary (curved arrow) . Also note the ovarian vessels (arrowheads) leading to the ovary. C, Positron emission tomography/CT image shows increased metabolic activity in the transposed ovary (arrow) corresponding to the corpus luteal cyst. Presence of this finding should not be misinterpreted as a metastatic deposit.


CT is an important modality when evaluating patients with neoplastic and nonneoplastic diseases. CT is not the preferred modality of choice for differential diagnosis of an adnexal mass, with the exception of a dermoid, because of its limited value in determining whether the mass is benign or malignant. In the preoperative staging of ovarian cancer, however, CT remains the most useful technique. CT has high sensitivity for detection of peritoneal implants in patients with ovarian cancer, especially when thinner slices (3 mm) and multiplanar images are used ( Figure 3-32 ). 10

FIGURE 3-32 Coronal reconstruction computed tomographic image of a patient with ovarian cancer shows multiple tumoral implants along the undersurface of the diaphragm (arrows), which is better appreciated on the coronal reconstruction image compared with the axial image.
The current role of CT in cervical cancer is mainly the staging of the advanced tumors and evaluating patients for tumor recurrence. CT performs similarly to MRI in overall staging, but MRI surpasses CT in tumor visualization and local extension. 11
When evaluating endometrial cancer, recent data show that a 16-slice CT scanner allows an accurate estimation of the local extent of the disease in patients with endometrial carcinoma by predicting the depth of myometrial invasion and the presence of cervical infiltration with rates similar to those reported for MRI. With the introduction of isotropic imaging and multiplanar reformations, images can be interpreted in perpendicular plane to the axis of the uterus, allowing more accurate evaluation of the local extent. 12
In addition to staging, CT is commonly used to evaluate tumor recurrence, assess tumor response to treatment, and plan radiation therapy. Other frequent indications for CT include acute and chronic pelvic pain, pelvic inflammatory disease, and equivocal findings on pelvic ultrasound.

Every female patient of reproductive age should be asked about the possibility of pregnancy before performing CT examination. CT has traditionally been avoided during pregnancy because of its ionizing radiation and its risk for teratogenesis and carcinogenesis. Other imaging modalities such as ultrasound or MRI should be considered first, unless benefits of performing the CT outweigh the risks of radiation. One of the most common indications for obtaining CT in pregnant patients is history of trauma when the risk for radiation to the fetus is outweighed by the benefits of preserving maternal health.
If iodinated contrast is given to the pregnant patient, free iodide has a theoretical potential to depress fetal/neonatal thyroid function. Infants who have undergone CT with contrast media in utero are screened for hypothyroidism during the first week of life. If iodinated contrast is given during the lactation period, many institutions recommend that the baby should not be breast-fed for 24 to 48 hours. However, because only trace amounts of iodinated contrast reach the neonate’s circulation, some authors suggest that it is likely not necessary to disrupt the breast-feeding process after the examination. 13
Intravenous contrast should not be given to patients with history of severe allergic reaction to the contrast agent and history of renal failure. If it is absolutely necessary to administer the contrast material, specific precautions should be taken to minimize the risks.

Technique Description
Adequate opacification of the small and large bowel by oral contrast agent is essential for most CT examinations of the pelvis. Oral contrast administration is recommended for adequate bowel opacification when evaluating the pelvis to avoid mistaking bowel loops for an abscess or a pelvic mass. Timing of the contrast administration is also important because it is important to opacify loops of small bowel in the pelvis. If the patient is scanned too early, the contrast might have not reached the loops of bowel in the pelvis. Conversely, if too much time elapses between contrast administration and image acquisition, most of the contrast will have moved out of the small bowel into the colon. A typical dose is 1000 to 1500 mL of iodine- or barium-containing contrast material, distributed over 1 hour. Compared with barium, iodinated oral contrast has a positive effect on peristalsis, resulting in somewhat faster transit times. Negative oral contrast could also be given; however, there is a risk for mistaking pelvic masses for undistended loops of bowel. Some institutions routinely administer rectal contrast, especially when there is a history of pelvic abscess or malignancy. Rectal contrast material should not be administered when there is active inflammatory bowel disease or a recent rectal anastomosis because perforation may occur.
CT of the pelvis is generally performed after administration of intravenous contrast media. The improved soft tissue contrast helps better define pelvic structures as a result of patterns of parenchymal opacification and helps differentiate pelvic lymph nodes from vessels. Additionally, concurrent CT of the abdomen is commonly performed, and presence of intravenous contrast aids in detection of distant metastasis. Typically 100 to 150 mL of low osmolar contrast media, 300 to 350 mg I/mL, is administered using a power injector at a rate 2 to 4 mL/s. The use of a saline bolus after the contrast administration compresses the contrast agent bolus and improves vessel enhancement in angiography. The introduction of multislice technology resulted in decreased scanning times, allowing coverage of the entire abdomen and pelvis during optimal enhancement. Various contrast delay times are acceptable. When dedicated CT of the pelvis is performed, a delay of 90 to 120 seconds is optimal for the venous enhancement, which helps differentiate pelvic lymph nodes from the vessels because pelvic vessels usually opacify after 90 seconds. Alternatively, when performed as a part of abdominopelvic study, a shorter delay time may be required. Delayed imaging after 3 to 5 minutes can be useful when evaluating the bladder and distal ureters. Helical CT offers the possibility to scan in craniocaudal and caudocranial directions. Both approaches are acceptable.

Pitfalls and Solutions
A common error is mistaking cystic masses in the pelvis for unopacified loops of small bowel ( Figures 3-33 through 3-35 ). A scan delay of at least 1 hour after oral contrast is administered is recommended to ensure adequate opacification of the bowel. Similarly, adequate opacification of small bowel is important to differentiate lymph nodes from bowel. Lymph nodes can occasionally be mistaken for vessels, and scanning the patient with a delay of 90 to 120 seconds is helpful to better define the vessels. Review of reformatted sagittal projections is often useful in identifying enlarged internal iliac and obturator chain nodes. A normal unenhanced bladder can be mistaken for a cystic ovarian mass. Delayed images will demonstrate contrast media filling the bladder.

FIGURE 3-33 Lymphocele in a 57-year-old patient with history of hysterectomy and lymph node dissection for endometrial carcinoma. A, Cystic structure near the pelvic sidewall (arrow) could be mistaken for an unopacified loop of bowel in the absence of oral contrast. B, Presence of oral contrast confirms that this structure is a separate entity from adjacent loops of bowel.

FIGURE 3-34 Computed tomography scan of a 68-year-old woman with history of ovarian cancer. Complex cystic metastasis (arrow) is well seen because of presence of oral contrast.

FIGURE 3-35 Pelvic inflammatory disease with pyosalpinx in a 26-year-old patient. Computed tomographic image through pelvis shows cystic tubular structure with thick enhancing walls (arrow) lateral to the uterus, which, in absence of oral contrast, could be mistaken for a loop of small bowel. Tuboovarian abscess is present to the right of the uterus (arrowhead) .

Multidetector Computed Tomography Technique
After the introduction of the first four-slice MDCT scanner in 1998, CT technology has advanced rapidly. The most important developments include increase in x-ray tube rotation speed and increase in the number of detectors. The major improvements of 16-slice CT compared with 4-slice CT are improved temporal resolution resulting from shorter gantry rotation time, considerably reduced scan acquisition times, and better spatial resolution owing to submillimeter section collimation. 14 Section collimation is determined by available detector configuration. The minimum section collimation for 4-, 16-, and 64-slice scanners ranges from 0.625 to 1 mm depending on the vendor.
Choosing section collimation is important because the thickness of the reconstructed data set must exceed it. Thus if diagnostic images are to be viewed on 2.5-mm-thick sections, then the acquisition collimation must not be more than 2.5 mm. In the abdomen and pelvis, reconstructed sections are reviewed 3 to 5 mm thickness. The originally required data set is usually retained for at least 24 hours to allow for special reformatting requirements.
In isotropic imaging, dimensions of each voxel in the data set are equal. The 1-mm collimation thickness used in four-slice MDCT does not fully match the in-plane resolution of approximately 0.5 to 0.7 mm. True isotropic resolution for routine applications had been achieved with introduction of 16-slice scanners because of their submillimeter collimation. These isotropic data allow reformatting of images in any desired plane having similar resolution to that of the original scanning plane yielding reformatted data sets of high quality. 15
One of the main advantages of MDCT is shorter acquisition times, which make it possible to scan the entire abdomen and pelvis in a single breath-hold. This reduces motion artifacts especially in uncooperative and critically ill patients. Acquiring images of the abdomen and pelvis may take 20 to 30 seconds using 4-slice CT machines, and the same acquisition may be completed in 5 to 8 seconds using 64-slice machines.
Pitch is the ratio between slice thickness and table speed. Pitch is high when the table moves a greater distance in relation to slice thickness. In a high-pitch setting, scanning times are shorter, the dose to the patient is lower, and the images are less detailed. In a low-pitch setting, scanning times are longer, the patient’s dose is higher, but the images are of better quality. In general, keeping the pitch between 1 and 2 provides a balance between coverage and resolution.
Tube voltage (kilovolt peak) determines the energy of x-ray beams. Variations in the tube voltage affect the image noise, contrast, and dose to the patient. If the tube voltage is decreased, patient dose is decreased, and image noise and image contrast are increased. Most CT studies are performed at 120 or 140 kVp. Higher settings (140 kVp) are advantageous for obese patients. For most other applications, a setting of 120 kVp will suffice. Lower settings (80 to 100 kVp) can be used to accentuate contrast enhancement of various structures because attenuation of iodine increases at these lower settings. When the tube voltage is decreased, it is important to increase milliampere-second settings to overcome the increase in image noise that occurs at lower kilovolt peak settings.
For postprocessing purposes, CT images should be acquired with the thinnest reasonable collimation. Secondary data sets can then be reconstructed using the original data set by creating images with 30% to 50% overlap, which will serve as source for displaying multiplanar two-dimensional (2D) images as well as 3D images. Processing techniques that result in 2D images are coronal, sagittal, and curved multiplanar reformations (MPRs). The key to optimizing image quality of MPR is to increase the reconstruction thickness to several (2 to 3) millimeters when manipulating the source data. A limitation to the routine use of nonaxial reformations is the time taken to perform MPR images. Recent software upgrades to multislice CT machines allow automatic reformations in any preselected plane to be performed within a few minutes at the end of the examination. Evaluation of MPR images is particularly helpful in the pelvis for assessing the extent of the disease such as seen in gynecologic malignancies ( Figure 3-36 ).

FIGURE 3-36 Sagittal reconstruction computed tomographic image of a 38-year-old woman shows the extent of hypoenhancing cervical mass (arrows) that is infiltrating into the bladder (arrowhead) .
The most commonly used image processing techniques that result in 3D images are shaded surface display, maximum intensity projection, minimum intensity projection, and volume-rendering technique. These 3D images have applications in the evaluation of skeletal structures and vascular anatomy.
Radiation protection is a critical concern for all CT examinations, especially in young adult females. Several scanning factors affect the radiation dose to the patient. One of the most effective methods of controlling the radiation dose is automated exposure control (AEC), which uses tube current modulation technique. AEC technique adjusts the radiation dose according to the patient’s attenuation while sustaining diagnostic image quality. Dose reduction techniques will be discussed in detail in Chapter 4 .


Recognizing the normal anatomic variants and normal postoperative CT appearances of female anatomy is essential.
Advances in MDCT technology, particularly the use of reformatted images and angiography, have significantly advanced the role of CT in imaging of female pelvis.
Appropriate timing of oral and intravenous contrast material administration improves diagnostic accuracy.
Precautions should be taken to minimize the risk of ionizing CT radiation.


1. Standring S., editor. Gray’s anatomy: the anatomical basis of clinical practice, ed 39, London: Elsevier/Churchill Livingstone, 2005.
2. Grossman J., Ricci Z.J., Rozenblit A., et al. Efficacy of contrast-enhanced CT in assessing the endometrium. AJR Am J Roentgenol . 2008;191(3):664-669.
3. Kaur H., Loyer E.M., Minami M., et al. Patterns of uterine enhancement with helical CT. Eur J Radiol . 1998;28(3):250-255.
4. Cohen H.L., Tice H.M., Mandel F.S. Ovarian volumes measured by US: bigger than we think. Radiology . 1990;177(1):189-192.
5. Borders R.J., Breiman R.S., Yeh B.M., et al. Computed tomography of corpus luteal cysts. J Comput Assist Tomogr . 2004;28(3):340-342.
6. Lee J.H., Jeong Y.K., Park J.K., et al. “Ovarian vascular pedicle” sign revealing organ of origin of a pelvic mass lesion on helical CT. AJR Am J Roentgenol . 2003;181(1):131-137.
7. Hiromura T., Nishioka T., Nishioka S., et al. Reflux in the left ovarian vein: analysis of MDCT findings in asymptomatic women. AJR Am J Roentgenol . 2004;183(5):1411-1415.
8. Vinnicombe S.J., Norman A.R., Nicolson V., et al. Normal pelvic lymph nodes: evaluation with CT after bipedal lymphangiography. Radiology . 1995;194(2):349-355.
9. Yassa N.A., Ryst E. Ovarian vein thrombosis: a common incidental finding in patients who have undergone total abdominal hysterectomy and bilateral salpingo-oophorectomy with retroperitoneal lymph node dissection. AJR Am J Roentgenol . 1999;172(1):45-47.
10. Pannu H.K., Horton K.M., Fishman E.K. Thin section dual-phase multidetector-row computed tomography detection of peritoneal metastases in gynecologic cancers. J Comput Assist Tomogr . 2003;27(3):333-340.
11. Hricak H., Gatsonis C., Coakley F.V., et al. Early invasive cervical cancer: CT and MR imaging in preoperative evaluation: ACRIN/GOG comparative study of diagnostic performance and interobserver variability. Radiology . 2007;245(2):491-498.
12. Tsili A.C., Tsampoulas C., Dalkalitsis N., et al. Local staging of endometrial carcinoma: role of multidetector CT. Eur Radiol . 2008;18(5):1043-1048.
13. Webb J.A., Thomsen H.S., Morcos S.K. Members of Contrast Media Safety Committee of European Society of Urogenital Radiology (ESUR). The use of iodinated and gadolinium contrast media during pregnancy and lactation. Eur Radiol . 2005;15(6):1234-1240.
14. Saini S. Multi-detector row CT: principles and practice for abdominal applications. Radiology . 2004;233(2):323-327.
15. Sandrasegaran K., Rydberg J., Akisik F., et al. Isotropic CT examination of abdomen and pelvis: diagnostic quality of reformat. Acad Radiol . 2006;13(11):1338-1343.

Suggested Readings

Foshager M., Walsh J. CT anatomy of female pelvis: a second look. Radiographics . 1994;14:51-66.
Langer J.E., Jacobs J.E. High-resolution computed tomography of the female pelvis: spectrum of normal appearances. Semin Roentgenol . 1996;31(4):267-278.
Chapter 4 Dose Reduction Techniques in Multidetector Computed Tomography Body Imaging

Lynne M. Hurwitz, Tracy A. Jaffe

Medical use of radiation has increased during the past decade and currently is estimated to account for approximately 50% of the radiation exposure to the U.S. population, with computed tomographic (CT) examinations accounting for the majority of this exposure. 1 The harm from radiation during diagnostic imaging studies is related primarily to the risk for cancer induction, which conservatively is determined using a nonthreshold linear relationship, such that the risk increases linearly with the radiation exposure. Determination of this risk takes into account the amount and type of radiation, age at exposure, total exposure, gender of the person, and organs that are exposed.
Acceptable image quality requires a certain amount of radiation exposure to the patient during multidetector computed tomographic (MDCT) examinations to ensure adequate signal-to-noise ratio (SNR) needed for confident diagnostic interpretation. MDCT imaging parameters can be adjusted depending on the specific indication of the study and body habitus of the patient to optimize image quality and reduce radiation dose exposure. In this chapter we discuss current techniques available to reduce radiation dose during MDCT body imaging.

Issues Pertinent to Radiation Dose for Multidetector Computed Tomography Studies
Radiation dose reduction must be performed within the context of maintenance of adequate image quality. When performing MDCT examinations the parameters of the examination should be chosen based on the type of the scanner, the size of the patient, anatomic region of interest, and the clinical question(s) to be answered. Parameters of the MDCT scanner that affect radiation include scanner geometry, tube current, tube voltage, scanning modes, scan length, collimation, table speed, gantry rotation time, and filters. Because of differences in scanners, including detector configuration and quantity of output of the x-ray beam, migration of protocols from one scanner to another often requires changing protocol parameters. Understanding the specifics of the scanner allows for appropriate selection of parameters to minimize radiation exposure while maintaining image quality.
Variations in body size and density will attenuate the x-ray beam differently. As a generalization, more radiation is needed to penetrate larger patients as a result of scatter of the primary beam from a larger body size. Although increases in tube current may be necessary in imaging obese patients, it has been shown that this modification does not increase the absorbed dose for internal organs. 2 In certain anatomic locations such as the pelvis, there is increased attenuation of the x-ray beam by the surrounding bones and higher amounts of radiation are needed in this region to maintain image quality comparable to that in the thorax. Additionally, imaging in anatomic locations that have intrinsically low contrast-to-noise ratios, such as the pelvis, require a higher SNR than in other anatomic locations, such as the chest.
The clinical question for which the scan is being performed needs to be considered in regard to the reconstruction algorithm, z-axis coverage, and the necessary SNR, all of which will determine the overall radiation exposure and parameter settings for the CT examination. For example, imaging of the abdomen and pelvis for renal calculi can be done with an overall lower radiation dose than for appendicitis imaging because of the intrinsically high contrast-to-noise ratio present between a small renal stone and the surrounding ureter ( Figure 4-1 ).

FIGURE 4-1 Axial computed tomographic (CT) images demonstrating difference in image quality for renal calculus versus appendicitis CT protocols. A, Acute abdomen CT protocol (450 mA) with appendicitis (arrow) B, Renal stone protocol CT (240 mA) with left renal calculus (arrow) . Increased noise does not degrade stone detection.

Strategies for Dose Reduction
The parameters of the CT examination can be altered in isolation; however, for effective dose reduction and maintenance of image quality, maximal effect will be achieved if the parameters are evaluated in relation to each other.

Tube Current
Current MDCT tubes can generate outputs up to 800 mA. A constant milliampere can be applied for evaluation of the abdomen and pelvis; however, software advances now allow for the milliampere to vary during the study while maintaining a constant noise level; this results in lower radiation exposure to the patient in the range of 20% to 35% for body imaging. 3, 4 This application, known as automatic tube current modulation (ATCM), can be performed in several manners depending on the individual CT scanner being used. Angular dose modulation occurs in the X and Y plane, reducing the milliampere in the anterior and posterior projections while maintaining a higher tube current along the lateral projection (compared with the anterior and posterior projections), which contributes more to the image noise; this reduction in dose is usually based on the previous 180 to 360 degrees of scanning. Modulation of dose can also occur along the z-axis of the patient, this type of modulation is based on the scout film and has the ability to reduce dose in anatomic areas of the body that attenuate the beam the least, such as the chest compared with the abdomen ( Figure 4-2 ). A combination of these two techniques is also available on many scanners.

FIGURE 4-2 Tube current modulation based on scout film for abdomen and pelvic computed tomography. Green line demonstrates change in milliamperes (y axis) during course of data acquisition (constant noise level) for the study.
ATCM software uses several input parameters for appropriate application. A user input value that describes the type of desired image quality (noise index, reference image acquisition, reference tube–current time product value, or reference standard deviation) is required. This value can vary depending on the body habitus, the indication of the study, and the reconstruction algorithm. Appropriate orientation of the patient with respect to the gantry is necessary for ATCM to work effectively. Miscentering the patient within the scanner can change the distance between the patient and the x-ray beam resulting in either magnification or minification of the scout film. This misrepresentation of the body can result in an inappropriate increase or decrease in the milliampere settings. 5, 6 The relationship of the beam to the patient also will change such that the beam will no longer be centered on the patient; rather, it may be located more anteriorly or posteriorly, resulting in a change to the absorbed dose to the more superficially located organs.

Tube Energy
The options on the CT scanner for the tube energy setting are more limited than those of milliampere, ranging from 80 to 140 kVp. Historically a setting of 140 kVp was used for routine clinical examinations to preserve tube life and to allow for optimal penetration of a wide variety of body sizes. Recent work has demonstrated that for a smaller body habitus, a reduced peak kilovoltage can be used on a routine basis to reduce radiation exposure while maintaining adequate image quality. Additionally, when administering intravenous contrast, reducing the peak kilovoltage will not only lower the overall radiation exposure but also can improve the image quality because imaging is performed closer to the k-edge of iodine (33 kV). 7 This has shown to be of particular benefit when performing dedicated MDCT angiograms.

Z-Axis Coverage
Reducing the z-axis coverage will linearly decrease the overall radiation exposure to the patient. For abdominal and pelvis imaging this is of considerable importance because imaging caudal to the pubic symphysis and cranial to the diaphragms has been shown to have minimal change in patient outcomes while increasing radiation exposure to patients. 8 Increasing the z-axis coverage above the diaphragms contributes to greater radiation exposure to the breasts, and this has been noted to be of particular concern for young female patients. 9

Scan Time
The acquisition time for a CT examination will vary depending on z-axis coverage (discussed previously), rotation time, pitch, and beam collimation (width). A faster gantry rotation time will decrease radiation exposure to the patient if all other parameters are held constant and may also result in fewer motion artifacts. Increasing the pitch can result in a lower radiation dose. This can lead, however, to generation of helical artifacts and reduction in SNR. Decreasing the pitch will increase the radiation dose; however, this is often needed for MDCT angiograms, for which interslice gaps need to be minimized.
Using a larger beam collimation (width) will increase the z-axis coverage per rotation. Because of the penumbra effect, a small portion of the x-ray beam is used during helical scanning to maintain consistent image quality to all detectors (known as overbeaming); the contribution of overbeaming to total radiation exposure can be minimized by using a wide beam width and/or higher pitch. Depending on the configuration of the scanner, wider beam widths may require using wider length detectors (i.e., scanning at 1.25 vs. 0.625 mm), which may limit one to a larger slice thickness width.

Filtration of the x-ray beam can allow for removal of photons that do not add to image quality but contribute to a patient’s radiation exposure. Bowtie filters shape the x-ray beam, and when matched to the field of view, the scatter of the beam outside the field of view is reduced. This has been shown to reduce the amount of radiation on the surface of the body. 10 Bismuth shields placed on the patient lower energy photons that do not add to image quality but increase radiation exposure to the patient. These have been shown to be of particular use in reducing the radiation filter dose to superficial organs such as the breast, thyroid, and eyes.
Postprocessing filters can also be applied to improve image quality and their use with lower radiation dose acquisition is unclear. 11

Reconstruction Algorithms
The prescribed reconstruction algorithm will determine how much noise is acceptable for the examination. Although current scanners can acquire isotropic data sets, more frequently data sets are viewed in reconstructed 3- to 5-mm-thick images. Thicker reconstructed data sets can accommodate more noise without degrading the image quality. For example, when a thick reconstruction algorithm is used (i.e., 5- vs. 1.25-mm-thick images), a lower radiation dose technique can be used (4×) to maintain a similar image quality. 12 Care should be taken, however, with this approach because partial volume averaging effects will be amplified. Currently, iterative reconstruction techniques are being investigated as an alternative mode of data reconstruction to filter back projection techniques to reduce radiation exposure during MDCT examinations.

Clinical Scenarios with Focus on Radiation Exposure

Case 1

A 32-year-old woman with new-onset right lower quadrant pain and elevated white blood cell count
CT request: Evaluate for appendicitis
Protocol options:
1. CT of the abdomen and pelvis
2. Ultrasound pelvis and right lower quadrant at point of pain
3. Magnetic resonance imaging (MRI) of the abdomen and pelvis
Assessment of acute-onset right lower quadrant pain in women can be challenging because of the complex pathologies that can occur in this region from gastrointestinal and genitourinary sources. Radiation dose is of particular concern in young women, and use of imaging techniques that do not use ionizing radiation is desirable. Abdominal and pelvic ultrasound can provide accurate and detailed information localized to the area of pain. Traditionally ultrasound has been the imaging modality of choice for assessment of ovarian and acute pelvic processes because it is readily available, has adequate spatial resolution, particularly when endovaginal ultrasound is used, and allows for real-time visualization. CT imaging is appropriate in this patient population when there is greater concern for gastrointestinal pathology related to the right lower quadrant pain. To reduce radiation, CT techniques should limit z-axis coverage, confine scanning to a single contrast phase, and use automatic dose modulation and an increased pitch value.
There are several patient populations who are at risk for repeated radiation exposure in the setting of recurrent abdominal pain. These include patients with renal colic, a diagnosis that is now routinely evaluated with CT, and young patients with Crohn’s disease. All effort to minimize radiation exposure in these patients should be undertaken. Alternative imaging, especially MRI in those with Crohn’s disease, may be warranted. MRI can also be used to accurately characterize the female genitourinary system because of its high contrast-to-noise ratio compared with CT and improved spatial resolution over CT and ultrasound.

Case 2

A 45-year-old woman status post radiation and chemotherapy treatment for cervical carcinoma ( Figure 4-3 ) presents for assessment of disease progression
CT request: Staging CT
Protocol options:
1. CT of the abdomen and pelvis
2. Whole body positron emission tomography (PET)/CT
3. MRI of the abdomen and pelvis

FIGURE 4-3 Cervical cancer. A, Axial computed tomographic (CT) image through the pelvis demonstrates a markedly enlarged cervix with necrosis (arrow) B, Coronal image of fused positron emission tomography/CT demonstrates a mass in the pelvis (arrow) and iliac adenopathy (arrowhead) .
Monitoring of tumor response to therapy is routinely performed with CT. With the exception of choriocarcinoma, gynecologic malignancies do not demonstrate hypervascularity, and imaging should be performed in a single phase of imaging timed to portal venous phase of contrast enhancement. As in routine imaging of the female patient, care should be made to optimize CT protocols for radiation exposure, including breast shielding. MRI delivers no ionizing radiation and should be used for disease surveillance when available; using T2-weighted images are preferred. The use of dynamic contrast enhanced T1-weighted images improves accuracy. 13 PET/CT is increasingly being used for gynecologic malignancy recurrence surveillance.

Case 3

A 45-year-old woman status post bowel resection with fever and elevated white blood cell count ( Figure 4-4 )
CT request: Rule out abscess
Protocol options:
1. CT of the abdomen and pelvis
2. MRI of the abdomen and pelvis
3. Ultrasound of the abdomen and pelvis

FIGURE 4-4 Abscess. Axial computed tomographic image after the administration of oral and intravenous contrast through the pelvis in a patient status post bowel resection. Contrast-filled bowel loops are separate from large fluid collection (white arrows) . A second fluid collection is seen posteriorly (black arrow) behind opacified loops of bowel.
The evaluation of the postoperative abdomen and pelvis should be performed with CT (with the notable exception of pregnancy). Both pneumoperitoneum and hemoperitoneum are less conspicuous on MRI and may be inadvertently overlooked, which can lead to a delay in diagnosis. If possible, CT should be performed with a positive oral contrast agent to increase conspicuity of fluid collections and luminal extravasation. The administration of intravenous contrast also increases conspicuity of fluid collections. MRI is of value in the identification of pelvic fistulas and determining the contents of fluid collections. Ultrasound imaging within the peritoneal cavity is of limited use secondary to overlying bowel gas and widespread scope of search.

Case 4

A 23-year-old woman 15 weeks pregnant with new-onset dyspnea
CT indication: Evaluate for pulmonary embolus
Protocol options:
1. Pulmonary embolus CT angiography (CTA)
2. Ultrasound deep venous thrombus study
3. Ventilation–perfusion (V/Q) scan
4. Chest radiograph
Imaging during pregnancy requires particular attention to imaging technique to reduce the radiation dose to the mother and the fetus. Maternal morbidity from a missed diagnosis can be fatal to the mother and fetus, and significant radiation exposure can also increase the mother’s risk for cancer induction and the fetus’ risk for cancer induction or neurologic detriment from the radiation exposure. Imaging modalities that use minimal or no radiation should be considered, including lower extremity ultrasound and chest radiographs, which can diagnose a deep venous thrombosis or provide an alternative cause for shortness of breath, such as pneumonia. If these fail to provide a diagnosis and further imaging is indicated, all efforts to reduce radiation exposure to the mother and fetus should be used. Radiation doses from both chest CTA and V/Q scans result in low fetal radiation exposure. 14 V/Q imaging will result in lower material breast-absorbed radiation dose compared with chest CTA. 9 In patients with a normal chest radiograph, a V/Q scan may provide overall better quality study compared with chest CTA as a result of the increased maternal vascular volumes that can result in suboptimal contrast opacification of the pulmonary arteries and hemodynamic changes that can also increase cardiac and respiratory motion artifacts during chest CTA examinations ( Figures 4-5 and 4-6 ).

FIGURE 4-5 Pulmonary embolus during pregnancy. A, Posteroanterior chest radiograph demonstrates normal lungs and pleural spaces with an enlarged cardiac silhouette. B, Ventilation–perfusion study demonstrates normal ventilation and multiple unmatched segmental perfusion defects.

FIGURE 4-6 Pulmonary CTA during pregnancy. Coronal and sagittal oblique views of the left and right pulmonary arterial system (arrows) demonstrate decreased enhancement of the lower lobe pulmonary arteries.
In the setting of trauma, CT is the first-line imaging modality for imaging the pregnant woman because preservation of maternal health is the greatest predictor of fetal health. Triage for CT is fast, as is the acquisition of CT images. CT is superior to MRI for the identification of small volumes of hemoperitoneum and pneumoperitoneum.


Radiation dose reduction must be performed within the context of maintenance of adequate image quality.
Current MDCT scanners vary in detector configuration, rotation speed, and quantity of output of the x-ray beam; as a result, migration of protocols from one scanner to another requires alterations in imaging parameters.
Automatic tube dose reduction software and bismuth breast shields should be used whenever possible.
Centering the patient and scanning only the area of interest contribute to decreasing radiation dose.
The peak kilovoltage can be decreased from 140 to 120 with improved image appearance quality when intravenous contrast is administered.
For the assessment of primary nontraumatic abdominal pain in the pregnant woman, ultrasound is considered the initial study of choice. In cases where ultrasound is nondiagnostic, MRI or CT may be performed for further evaluation. General trends in abdominal imaging show a preference for MRI in all three trimesters for the evaluation of nonurinary, nontraumatic abdominal pain during pregnancy, although there is continued debate over the appropriate imaging modality of choice in the setting of urinary symptomatology during the second and third trimesters.


1. Report No. 160: Ionizing radiation exposure of the population of the United States . Bethesda, MD: National Committee on Radiation Protection; 2009.
2. Schindera S.T., Nelson R.C., Lee E.R., et al. Abdominal multislice CT for obese patients: effect on image quality and radiation dose in a phantom study. Acad Radiol . 2007;14:486-494.
3. Tack D., De Maetelaer V., Gevenois P.A. Dose reduction in multidetector CT using attenuation-based online tube current modulation. AJR Am J Roentgenol . 2003;181:331-334.
4. Mulkens T.H., Bellinck P., Baeyert M., et al. Use of an automatic exposure control mechanism for dose optimization in multidetector row CT examinations: clinical evaluation. Radiology . 2005;237:213-223.
5. Matsubara K., Koshida K., Ichikawa K., et al. Misoperation of CT automatic tube current modulation systems with inappropriate patient centering: phantom studies. AJR Am J Roentgenol . 2009;192:862-865.
6. Li J., Udayasankar U.K., Toth T.L., et al. Automatic patient centering for MDCT: effect on radiation dose. AJR Am J Roentgenol . 2007;188:547-552.
7. Nakayama Y., Awai K., Funama Y., et al. Lower tube voltage reduces contrast material and radiation doses on 16-MDCT aortography. AJR Am J Roentgenol . 2006;187:W490-W497.
8. Kalra M.K., Maher M.M., Toth T.L., et al. Radiation from “extra” images acquired with abdominal and/or pelvic CT: effect of automatic tube current modulation. Radiology . 2004;232:409-414.
9. Hurwitz L.M., Yoshizumi T.T., Reiman R.E., et al. Radiation dose to the female breast from 16-MDCT body protocols. AJR Am J Roentgenol . 2006;186:1718-1722.
10. Toth T.L., Cesmeli E., Ikhlef A., et al. GE Healthcare, Waukesha WI, USA, GE Healthcare, Hino Japan. Image quality and dose optimization using novel x-ray source filters tailored to patient size. Proceedings of SPIE. Physics of Medical Imaging . 2005;5745-34:1-9.
11. Kalra M.K., Maher M.M., Blake M.A., et al. Detection and characterization of lesions on low-radiation-dose abdominal CT images post processed with noise reduction filters. Radiology . 2004;232:791-797.
12. Kanal K.M., Stewart B.K., Kolokythas O., et al. Impact of operator-selected image noise index and reconstruction slice thickness on patient radiation dose in 64-MDCT. AJR Am J Roentgenol . 2007;189:219-225.
13. Kaur H., Silverman P.M., Iyer R.B., et al. Diagnosis, staging, and surveillance of cervical carcinoma. AJR Am J Roentgenol . 2003;180:1621-1631.
14. Pahade J.K., Litmanovich D., Pedrosa I., et al. Imaging pregnant patients with suspected pulmonary embolus: what the radiologist needs to know. Radiographics . 2009;29:639-654.

Suggested Readings

Kalra M.K., Maher M.M., Toth T.L., et al. Strategies for CT radiation dose optimization. Radiology . 2004;230:619-628.
Saini S. Multi-detector row CT: principles and practice for abdominal applications. Radiology . 2004;233:323-327.
Part Three
Magnetic Resonance Imaging
Chapter 5 Magnetic Resonance Imaging of the Female Pelvis
Technique, Anatomy, and Pitfalls

Julia R. Fielding, Alfred Llave
Magnetic resonance imaging (MRI) offers infinite gray scale resolution allowing for improved tissue characterization and multiplanar capabilities. This facilitates the diagnosis of a variety of benign and malignant conditions involving the female pelvis without the use of ionizing radiation. Common indications for MRI of the pelvis are pain, infertility, adnexal mass characterization, incontinence, and cancer staging, including assessment of posttherapeutic response and complications. There is increasing use of MRI for the evaluation of congenital anomalies, lymphadenopathy, and pain in pregnant patients during the second and third trimesters. The goal of this chapter is to provide an overview of basic MRI techniques and normal gynecologic anatomy, illustrative conditions, and common pitfalls.

Technical Requirements
Ideally, scanners with field strengths of at least 1.0 Tesla should be used to perform a diagnostic pelvic MRI study. Current standard of care for pelvic MRI includes the use of a multicoil array ( Figure 5-1 ), which increases signal-to-noise ratio and provides better anatomic resolution. The multicoil array consists of several adjacent receiver coils arranged in a belt, which when secured around the pelvis act as individually separate receivers of signal transmitted from the pelvic organs. The signals are acquired simultaneously with spatially encoded information, which are then summed before processing into an image. This improved signal acquisition allows for decreased scan time and improved image resolution compared with use of a standard whole-body magnet alone.

FIGURE 5-1 Multicoil array. Several adjacent coils act as receivers of signal, which are summed before processing into an image. Radiofrequency pulses are transmitted from the body coil.
Limitations in the use of multicoil array include misregistration artifact along the phase encoding direction, which is introduced by respiratory motion particularly when not properly secured or when applied over clothing. In some instances, compression bands may be used to minimize abdominal wall motion. Near-field bright artifact ( Figure 5-2 ) occurs with large amounts of subcutaneous fat, which accentuates motion artifact. This can be reduced by the application of saturation bands (5 cm in thickness) along the anterior abdominal wall fat with the position determined by the sagittal scout image ( Figure 5-3 ). The time required to add these bands precludes the use of ultrafast sequences, such as T2-weighted half-Fourier rapid acquisition with relaxation enhancement (RARE). Because signal-to-noise ratio is inversely related to the radius of the individual coils, images acquired from obese or pregnant patients and those with large pelvic masses extending into the abdomen demonstrate signal loss centrally. These patients are better imaged directly using the whole-body resonator alone ( Figure 5-4 ). A summary of common artifacts and remedies is presented in Table 5-1 .

FIGURE 5-2 Near-field artifact. Bright areas resulting from coil resting on abundant subcutaneous fat anteriorly and posteriorly. Radiofrequency penetration to central pelvis is limited secondary to body size.

FIGURE 5-3 Saturation band. Gray rectangle represents optimal location of saturation band over the anterior abdominal wall subcutaneous fat. This decreases motion and near-field artifact.

FIGURE 5-4 Loss of signal centrally occurs in individuals with large abdominal girths such as in pregnancy, obesity, or large masses as seen in this example. Depth of pulse penetration is equal to the radius of each superficial coil. Use of the main body resonator alone often improves image quality.
TABLE 5-1 Common Artifacts Artifacts Cause Remedy Double-contour artifact and blurring Motion (breathing, peristalsis, movement during examination) Fast imaging, properly secured multicoil array, respiratory coaching, compression bands, proper fasting and voiding, saturation bands, and fat suppression Near-field bright artifacts Large amount of subcutaneous fat Saturation bands, fat suppression Loss of signal in center of field of view Obesity, large masses, pregnancy or ascites, high magnet strength (3 Tesla) Image directly with body coil Nonuniform fat saturation Magnetic field inhomogeneity Optimization of field homogeneity, center anatomy within bore Magnetic susceptibility artifact Metal implants Use of spin echo sequence with 180-degree refocusing RF pulse, TSE with long echo trains Truncation artifacts Large amount of subcutaneous fat Fat suppression or inversion recovery False bladder wall thickening Chemical shift Change in direction of frequency encoding gradient, fat suppression
TSE, Spin echo; RF, radiofrequency.
Endorectal and endovaginal coils ( Figure 5-5 ) are used for higher resolution of the rectum, vagina, urethra, and periurethral anatomy. Both an endovaginal (Phillips, Andover, MA) and an endorectal coil (Siemens, Malvern, PA) are commercially available. Similar to the multicoil array coil, these coils operate as a receiver of signal from surrounding pelvic tissue before information is processed into an image. A disadvantage in the use of endoluminal coils is the limited field of view, where often the entire uterus cannot be seen or large pelvic masses are incompletely evaluated. These coils can be combined with the use of a multicoil array to obtain higher-resolution images of both the luminal structure and the adjacent pelvic organs. In general radiology practice, the most common use of the combined coils is for evaluation of the prostate gland.

FIGURE 5-5 Endorectal coil. Balloon has to be properly placed and inflated to minimize artifact related to motion.

Patient Preparation
A thorough gynecologic history, including the presence of pelvic devices and foreign bodies, should be obtained. Intrauterine devices ( Figure 5-6 ) and sterilization clips can safely go into the magnet with acceptable minimal artifact produced. Artificial joints are also safe; however, these result in more significant artifacts, particularly of the pelvic floor. As is true for all other MRI scans, nontitanium aneurysm clips, metal near the optic nerve, pacemakers, implantable defibrillators, and cochlear implants remain contraindications to MRI ( Box 5-1 ).

FIGURE 5-6 Intrauterine device (IUD) in uterus. High-resolution T2-weighted images demonstrate IUD in the appropriate position with no significant artifact. Uterus is retroflexed.

BOX 5-1 Contraindications to Magnetic Resonance Imaging

Cardiac pacemakers and defibrillators
Nontitanium aneurysm clips
Drug infusion pumps
Implantable defibrillators
Implanted electronic stimulators
Cochlear implants
Metal near the optic nerve
Patients should be requested to fast for 6 hours before imaging to minimize motion artifact from intestinal peristalsis. The use of glucagon has fallen out of favor because of increased cost and occasional paradoxical patient reactions. In recent years, its benefit has been further reduced as a result of improvement in scanning speeds, rendering image acquisition less motion sensitive. Voiding within 30 minutes before scanning is advised to minimize anatomic distortion artifacts related to bladder peristalsis and patient movement caused by discomfort. Compression bands can be used to decrease artifact related to bowel motion and respiration, although the latter is less problematic in imaging of the pelvis. Gastrointestinal contrast agents are not routinely used for pelvic examinations. Although some centers advocate the use of vaginal or rectal gels to help characterize surface abnormalities such as in endometriosis, this is not necessary for most examinations. Tampons are not used to evaluate the vaginal canal and can often alter the normal anatomy and be mistaken for pathology.
Because scans may last as long as 30 to 45 minutes, the patient is positioned supine, with legs supported under a bolster, to make the examination more comfortable. Earplugs are provided to the patient to help minimize noise related to some sequences.

Pulse Sequences and Imaging Planes
Correlating the various appearances of tissues on T1- and T2-weighted images, before and after contrast images, and without and with fat suppression allows for the evaluation of pelvic anatomy and disease. T2-weighted images best demonstrate the internal anatomy of pelvic organs because of inherent differences in the water content of various tissue types, and are helpful in assessing for pathology depicted by relative increase in water content. 1 - 3 , 5 - 10 This aids in the evaluation of cystic masses, free fluid, inflammation, necrosis, and fistulas. Because fat is bright on T1- and T2-weighted images, fat suppression is useful to make fluid more conspicuous. Compared with T2-weighted images, T1-weighted images are not as useful for the demonstration of the internal anatomy of the pelvic organs, but are excellent for definition of anatomic boundaries between soft tissues and fat planes, which are intermediate to dark and bright in signal, respectively. Fat suppression and in-phase and out-of-phase imaging techniques are useful in identifying certain fat-containing pelvic pathologies such as dermoids and differentiating them from hemorrhage-containing structures such as hemorrhagic cysts or endometriomas ( Figures 5-7 and 5-8 ). Certain pelvic disease states involving increased fluid such as necrosis and cystic and phlegmonous changes are low in signal, which become more identifiable in the background of bright fat. Calcifications are low in signal in both T1- and T2-weighted imaging, and fibrosis is intermediate on T1-weighted imaging and low on T2-weighted imaging.

FIGURE 5-7 Endometrioma. A, T1-weighted image demonstrates a well-circumscribed high signal mass in the left ovary. B, T1-weighted fat saturation image demonstrates persistent high signal signifying lack of fat within the mass. C, There is loss of signal on T2-weighted images. This is in contradistinction to dermoids, which would remain high in signal on T2-weighted images reflecting fat content (see Figure 5-8 ).

FIGURE 5-8 Dermoid tumor. A, Axial T1-weighted image demonstrates a high-signal mass with internal intermediate signal. B, Axial T1-weighted out-of-phase image shows peripheral rim of low signal, “India ink” effect, and decrease in signal of internal component consistent with fat in the mass with intravoxel water and fat within the darker internal component. C, Axial T1-weighted three-dimensional gradient echo image postcontrast with fat saturation demonstrates decreased signal of the mass, confirming fat content. D, Axial T2-weighted image shows persistent high signal of the mass comparable with subcutaneous fat.
Intravenous gadolinium-based contrast shortens T1, rendering vascular structures, inflammation, and neovascularity higher in signal. The timing of dynamic images helps to further characterize disease—specifically whether they are hypervascular, late enhancing, or non enhancing, suggesting angiogenesis, fibrosis, or necrosis, respectively. T2-weighted images are more helpful to characterize pelvic anatomy compared with postcontrast T1-weighted images. 1 When intravenous contrast is used, however, fat saturation is used to improve conspicuity of enhancement in the darker background of suppressed fat ( Figure 5-9) .

FIGURE 5-9 Lymphadenopathy. Fat saturation techniques are used identify an enlarged left pelvic sidewall lymph node (arrow) .
The first images of a study include a localizer sequence consisting of rapidly acquired coronal, sagittal, and axial T2-weighted images of the entire abdomen and pelvis. This is obtained to optimize positioning of the multicoil array over the area of interest. It is imperative that the array be positioned so that the anterior and posterior coils are at the same level. The large field of view of the initial scout also allows for cursory assessment of the kidneys and uterus, and detection of urinary tract obstruction ( Figure 5-10 ). The sagittal scout image is used to assess the position of the uterus, check for proper placement of endocavity coils, and for strategic positioning of the presaturation bands over the anterior and posterior abdominal adipose tissue. Finally, evaluation of the three planes allows for optimal planning of the subsequent smaller field-of-view series.

FIGURE 5-10 T2-coronal scout image. A, Value of the coronal half-acquisition rapid acquisition with relaxation enhancement scout image. Note pelvic kidney (arrow) and absence of kidneys in renal fossae. B, T2-coronal scout image in a different patient. Bilateral hydronephrosis (arrows) in a patient with cervical carcinoma with local invasion.
Multiplanar T2-weighted images are then obtained to provide excellent contrast resolution for the depiction of the uterine, ovarian, and cervix zonal anatomy, fluid-filled structures present in the female pelvis and free fluid. Because T2-weighted images are superior compared with the majority of other pulse sequences in depicting anatomic features and pathology in the female pelvis, sagittal and axial T2-weighted images should be obtained early in the examination when patient cooperation is optimal. Either free-breathing RARE or rapid breath-hold, such as half-acquisition RARE, sequences may be used. T2-weighted RARE images yield the highest spatial resolution but require approximately 3 minutes to obtain. Because the axial images are of paramount importance, T2-weighted RARE images should be obtained in this plane. In cooperative patients, rapid breath-hold sagittal and coronal T2-weighted sequences, using half-acquisition RARE, yield high-quality images in less than 25 seconds ( Figure 5-11 , A ). These T2-weighted sequences are often sufficient for benign conditions, such adenomyosis, mullerian anomalies, and in the general evaluation of fibroids. Although customarily used for T1-weighted imaging, fat saturation techniques with T2-weighted images can be used to demonstrate pathology, which is particularly helpful when intravenous contrast cannot be used. Fat signal can be nulled by using short time inversion recovery (STIR) or spectral adiabatic inversion recovery techniques resulting in better depiction of high signal related to inflammation or fluid ( Figure 5-11 , B ). The latter technique more selectively suppresses the fat signal and is less sensitive to motion, resulting in a better signal-to-noise ratio. 2 High-resolution images may also be obtained using isotropic fat-saturated proton density–weighted three-dimensional (3D)-RARE sequences with a 3-Tesla magnet ( Figure 5-11 , C ). Because the pelvis is less affected by motion related to diaphragmatic motion compared with imaging of the abdomen, free-breathing sequences are often sufficient, especially with the use of saturation bands.

FIGURE 5-11 A, Half-acquisition rapid acquisition with relaxation enhancement (RARE) image demonstrating normal ovaries (arrows) . High signal follicles are identifiable in virtually all women of menstrual age. B, Spectral adiabatic inversion recovery image shows fat saturation that highlights the ovaries (arrows) and uterine serosa without sacrificing signal. C, Isotropic fat saturation proton density–weighted three-dimensional RARE sequence. Note higher resolution of image detail with some loss of contrast.
Similar to T2-weighted sequences, T1-weighted image acquisition may be tailored according to the level of patient cooperation. As a result of repeated imaging, spin-echo sequences provide the highest resolution with free breathing but also require longer acquisition times. In cooperative patients the shorter time-requiring RARE pulse sequences provide good images. Slices are obtained with a single breath-hold, allowing limited coverage and repeated imaging. Both techniques can be combined with fat suppression. Shorter scan times are possible using two-dimensional (2D) spoiled gradient echo sequences, which also allow performance of in- and opposed-phase imaging. Adding an inversion pulse before the T1 gradient echo sequence increases the signal-to-noise ratio.
T1-weighted 3D gradient echo images yield thin-slice images usually 3.0 to 3.5 mm in thickness, allowing for thinner slice interpolation and vascular reconstructions. Although this sequence provides a lower signal-to-noise ratio compared with the previously described T1 sequences, the diagnostic information obtained after intravenous administration of gadolinium-based contrast agents (0.1 mg/kg body weight) is critical for the assessment of adnexal masses and staging of cancer. With gradient echo sequences, additional diagnostic information may be obtained by in-phase and out-of-phase imaging characterizing tissue with intravoxel water and fat (see Figure 5-8 ).
Contrast agents, usually gadolinium chelates, are best administered using a power injector at a rate of 2 mL/s followed by a saline flush. Arterial phase images are obtained using one of several techniques. A simple method uses a standard time of 25 seconds after injection to begin imaging. Patient variables such as cardiovascular status, weight, location, and size of intravenous catheter may affect the rate of arterial enhancement. A more customized strategy uses a test bolus of 2 mL of intravenous contrast hand injected at 2 mL/s. Region-of-interest monitoring above the aortic bifurcation is performed, which generates a time-density curve optimizing timing for the main bolus administered using a power injector. Most scanners are also equipped with a bolus tracking system that will automatically begin acquisition when the contrast bolus reaches the region of interest. Once the optimal delay time is determined, images are usually obtained during a breath-hold of approximately 15 to 20 seconds. Verbal or visual cues should be given to the patient to diminish breathing artifact. In healthy patients, optimal contrast between tumor and myometrium occurs between 2 and 3 minutes, and maximal endocervical mucosa enhancement occurs between 4 and 5 minutes. In most clinical cases, arterial phase images are obtained followed by delayed images performed 60 and 120 seconds from the time of injection. Data from the arterial phase scans can be used to reconstruct angiograms ( Figure 5-12 , A ) for either arterial embolization evaluation or surgical planning for tumor removal. Venograms can also be reconstructed ( Figure 5-12 , B ).

FIGURE 5-12 A, Magnetic resonance angiography. Coronal acquisition using three-dimensional gradient echo pulse sequence followed by maximum intensity projection. Optimum delay is approximately 25 seconds. B, Magnetic resonance venogram. Images can be obtained after a magnetic resonance angiography or in this case using a saturation band over the superior aspect of the body to block arterial inflow and a slightly longer delay.
Evaluation of vascular anatomy may also be performed without the use of intravenous contrast, measuring the motion of blood to visualize either arterial or venous flow. More commonly used techniques are the time-of-flight and phase-contrast magnetic resonance angiography techniques. As with contrast-enhanced magnetic resonance angiography, data obtained can be used to reconstruct maximum intensity projection models. Compared with contrast-enhanced magnetic resonance angiography, these noncontrast techniques are more sensitive to flow-related artifacts.
For detailed evaluation of small structures such as the vagina or urethra, small field-of-view images using high-resolution fast spin echo sequences are obtained. In evaluating fetal anomalies, multiplanar half-acquisition RARE or similar sequences with large fields of view are obtained, yielding snapshot images of the fetus.
The standard imaging planes used for imaging the pelvis are axial, sagittal, and coronal. As mentioned previously, large field-of-view coronal T2-weighted images are used as a scout image, along with the midsagittal scout view to be used as a guide for the application of saturation bands. The anatomy of the uterus and ovaries is best demonstrated in the axial plane. This plane is particularly important in the assessment of the parametrium and in the identification of pelvic lymphadenopathy. Therefore because the axial plane is most critical for diagnosis, it should be obtained first, especially in a patient who may tire easily or be uncooperative. For greater anatomic detail, the plane of acquisition may be adjusted along the long axis of the organ of interest based on the midsagittal scout.
The sagittal plane best depicts the zonal anatomy and position of the uterus and cervix. Anatomic relationships between pelvic compartment organs, particularly when tumor extension is being evaluated, are well demonstrated on this imaging plane. The coronal plane is complementary to the other planes and is particularly important in the assessment of pelvic lymphadenopathy ( Figure 5-13 ) and adnexal structures. A summary of optimal scanning planes is provided in Table 5-2 . A standard imaging protocol for the pelvis is provided in Table 5-3 .

FIGURE 5-13 Pelvic lymphadenopathy. Axial (A) and coronal (B) T2-weighted images demonstrate enlarged internal iliac lymph nodes from metastatic cervical cancer (arrows). Abnormal lymph nodes are often of increased signal compared with adjacent vessels and muscles on T2-weighted images. Short axis diameter greater than 1 cm is abnormal.
TABLE 5-2 Optimal Planes for Scanning Sagittal Uterus and cervix position, endometrium thickness, pelvic floor and relationship of pelvic compartment structures, cul-de-sac, presacral space Coronal Pelvic floor, uterine developmental anomalies, lymphadenopathy, adnexae Axial Lymphadenopathy, parametrium, vagina, cervix, uterus, ovaries Organoaxial Detailed evaluation of uterus, cervix, or vagina, parametrium

TABLE 5-3 Pelvic Magnetic Resonance Imaging Protocol at 1.5 Tesla With Body Array Coil

Normal Anatomy, Magnetic Resonance Imaging Appearances, and Common Differential Considerations


The vagina, derived from the Latin word meaning sheath , is a fibromuscular canal approximately 10 cm long, which extends from the cervix to the vulva. With its functions as a copulatory organ, outlet for menstruation, and birthing canal, it is composed mainly of thin layers of muscle and elastic tissue. It consists of an inner mucosal layer composed of nonkeratinized stratified squamous epithelium, an intermediate muscular layer with a bihelical arrangement of fibers, and an outer adventitial layer known as the paravaginal fascia, containing both dense and loose connective tissue and rich in blood vessels and nerves. Because of its pliable nature, its morphology is influenced both by its supporting ligaments and neighboring structures. Wider in the transverse plane, its anterior and posterior walls are approximated by the urinary bladder and urethra anteriorly and by the rectum posteriorly. With the impressions of these organs, along with the lateral wall support provided by the arcus tendineus, the lumen of the vagina assumes a butterfly or H configuration in the axial plane throughout most of its length. The uppermost aspect of the anterior vaginal wall is formed by the cervix, making it shorter than the posterior wall by approximately 3 cm. Reflections of the vaginal wall surrounding the outer cervix form the anterior, posterior, and the lateral fornices, which later form the upper margins of the lateral sulci. The cardinal and uterosacral ligaments along with the endopelvic fascia that support the cervix also provide support for the upper vagina. More inferiorly the vagina becomes constricted at the introitus, where it is fused with the urethra anteriorly and the perineal body posteriorly. The lateral support is provided by the fibers of Luschka of the pubococcygeus of the levator ani muscle group. Only the posterior vaginal fornix is covered by serosa, forming part of the cul-de-sac. The remainder of the organ is surrounded by fatty connective tissue called the paravaginal fascia. This richly vascularized connective tissue extends anteriorly to surround the urethra. Destruction of this fascia is associated with urinary incontinence.
The vagina is supplied by a rich network of vessels primarily from the descending branch of the uterine artery and branches of the internal iliac arteries, including the vaginal, rectal, vesical, and pudendal arteries. These branch vessels traverse its lateral wall and anastomose at the midline to supply the anterior and posterior vaginal walls. The pudendal artery provides branches to the inferior vagina. The posterior vaginal wall is also partly perfused from hemorrhoidal branches. The distal vagina is drained by the inguinal and anorectal lymph nodes, whereas the middle and upper third vagina drains into the paraaortic, internal and external iliac, sacral, and obturator group of lymph nodes.

Imaging Appearances
The vaginal canal is best evaluated using T2-weighted images in the axial plane, depicting its zonal anatomy and relationship to adjacent structures, including the paravaginal fascia. On T2-weighted images ( Figure 5-14 , A ), the low signal muscular layer with the closely apposed submucosa is clearly seen between areas of high signal: the mucosal epithelium and mucous secretions centrally, and the richly vascularized adventitia surrounding the vaginal muscular wall. 3 On the axial plane, these layers resemble a butterfly or H-shaped structure. On T1-weighted images, the vagina and adjacent tissues demonstrate intermediate signal and are therefore difficult to separate. After contrast enhancement and with the use of fat saturation, the mucosa and its secretions are low in signal, whereas the more vascular muscularis and adventitia are rendered higher in signal ( Figure 5-14 , B and C ). Sagittal images are useful for demonstrating the anterior and posterior vaginal fornices, as well as the pouch of Douglas, which is formed by the serosa covering the posterior aspect of the upper vagina and the anterior rectal wall ( Figure 5-15 ). The use of an endovaginal or endorectal coil can produce high-resolution images, providing excellent detail of perivaginal anatomy at the expense of a smaller field of view with the potential for patient discomfort.

FIGURE 5-14 Normal vagina. A, Axial T2-weighted image demonstrates the slightly higher signal mucosa (black arrow) , low signal muscular layer (white arrow) , surrounded by the high-signal paravaginal venous plexus. B, Axial postcontrast T1-weighted image with fat saturation. Central intermediate to low signal corresponds to the coapted mucosal layers with secretions (thin arrows) . The surrounding high signal corresponds to the muscularis and paravaginal venous plexus, which both enhance (thick arrows) C, Axial T1-weighted out-of-phase image demonstrates no significant difference in signal with the vaginal wall, perivaginal plexus, urethra, rectum, and pelvic floor muscles.

FIGURE 5-15 Normal vagina. Sagittal T2-weighted images demonstrate low signal vaginal wall (thin arrows) , as well as anterior and posterior fornices ( black and white arrows ). Note that the cervix forms the proximal third of the anterior vaginal wall. Small amount of free fluid is noted in the posterior cul-de-sac (star) . The uterus is anteflexed.
During reproductive ages, the T2-weighted appearance of the vagina is influenced by hormonal variations, 3, 4 which influence water content of structures. The inner zone, containing mucosa and secretions, is thicker during midcycle when estrogen levels are higher. The vaginal wall, however, is also relatively higher in signal during midcycle, resulting in less contrast to differentiate these structures. With low estrogen states as occurring during early proliferative and late secretory phases, the vaginal wall is lower in signal compared with the inherently high signal of the mucosa, allowing for better characterization. Similar low estrogen states that occur during the premenarche and postmenopausal periods result in the vaginal wall appearing low in signal with a much thinner central high signal mucosal stripe ( Figure 5-16 ).

FIGURE 5-16 Postmenopausal vagina. Sagittal T2-weighted image demonstrates thinning of the mucosa, which is not visible (C) . There is also decreased paravaginal tissue (arrows) . Free fluid indicated by star .

Common Differential Considerations
A variety of cystic structures are associated with or close to the vagina, which demonstrate high signal on T2-weighted images and variable signal on T1-weighted images, depending on the amount of mucin or the presence of inflammatory debris. Wall enhancement and thickening occur when these cysts are complicated by abscess formation. Gartner’s duct cysts are remnants of wolffian ducts, which are found in the upper anterolateral part of the vagina, and are usually less than 2 cm and asymptomatic. When enlarged, they may lead to urethral or ureteric obstruction and cause pelvic pain. Bartholin’s glands are located in the labia minora and have ducts that empty along the sides of the vaginal orifice. When they become obstructed, it leads to cyst formation ( Figure 5-17 ), which may become secondarily infected, leading to abscess formation. The urethra, closely abutting the lower vagina anteriorly, may present with diverticula. The anatomic relationships of these outpouchings may be better evaluated using small field-of-view high-resolution images in an axial plane oriented to the urethra ( Figure 5-18 ), allowing for better definition of its neck and possible complicating lithiasis. Suspected enterovaginal, rectovaginal, and cystovaginal fistulas are best evaluated with multiplanar high-resolution images and the use of endoluminal gels along with fat saturation techniques. Evaluation of suspected endometriosis involving the vaginal wall can also be facilitated with these gels. Presence of a vaginal tampon often distorts the vaginal anatomy and can mimic a mass ( Figure 5-19 ). In patients who have undergone hysterectomy, the residual vaginal cuff should appear smooth.

FIGURE 5-17 Bartholin’s cyst. Axial and parasagittal T2-weighted images (A and B) and axial T1-weighted postcontrast image (C) demonstrate a thin-walled cystic mass adjacent to the lower vagina, which is of fluid signal and does not enhance (arrow) . It is located inferior to the pubic ramus, which helps to differentiate it from a Gartner’s duct cyst, which is typically located above the inferior pubic ramus.

FIGURE 5-18 Urethral diverticulum. High-resolution T2-weighted axial (A) and sagittal (B) images show a crescent-shaped fluid collection surrounding the urethra (D) . Note debris layering in the dependent aspect. Diverticular stones can sometimes be found. Low intensity fibroids arise from the myometrium (F) .

FIGURE 5-19 Tampon. Sagittal T2-weighted image demonstrates intravaginal tampon with tip within the anterior fornix (arrow) . Distortion may simulate a vaginal or cervical mass.


The uterus, the Latin word for womb , is a pear-shaped, hollow muscular organ that receives the fallopian tubes. It is approximately 7.5 cm long and 5 cm wide in a nulliparous woman, and 5 cm long and 2 cm wide after menopause. It can be subdivided into three segments: the body, the isthmus, and the cervix ( Figure 5-20 ). The body, also known as the corpus , is the largest segment, comprising two thirds of the uterus. The fundus is the dome-shaped portion of the body located above the openings of the fallopian tubes. The body tapers inferiorly to a focal narrowing, the isthmus, which is also known as the lower uterine segment during pregnancy. The cervix, discussed separately, is the lowest part of the uterus, which protrudes into the upper vagina.

FIGURE 5-20 A through C, Illustrations depicting the anatomy of the female pelvis.
The body and fundus of the uterus are histologically identical and consist of three layers: the perimetrium, the myometrium, and the endometrium. The perimetrium is composed of mesothelium and loose connective tissue, which covers most of the uterus. This covering forms part of the vesicovaginal pouch anteriorly and the rectouterine pouch of Douglas, which extends inferiorly to the level of the posterior vaginal fornix and extends upward to cover the rectum anteriorly. The anterior surface of the lower uterus is surrounded by adventitia, whereas the broad ligament covers the lateral surface of the uterine corpus and cervix. The myometrium is the thickest layer and is composed of three ill-defined fibromuscular layers. The smooth muscle bundles in the outer and inner layers are oriented along the long axis of the uterus. The middle layer is the thickest layer with its muscle bundles oriented in a spiral fashion. Numerous large veins and lymphatics are present in this layer. The innermost region of the uterus is the endometrium, which is composed of epithelial cells, glandular elements, blood vessels, and collagenous connective tissue. This layer consists of two zones: the basalis and the functionalis. The basalis layer is adjacent to the myometrium and is rich in collagen. The functionalis layer is variable in thickness as it responds to hormonal stimulation of the menstrual cycle.
The uterus is supported by fascial condensations called ligaments. Lateral support is provided by the broad ligaments, which are peritoneal folds that drape over the lateral surface of the uterine body and fallopian tubes. The condensation of the connective tissues at the lower margins of the broad ligaments forms the cardinal ligaments that blend posteriorly with the uterosacral ligaments. The latter are formed by fibers extending from the sacrum that course around the rectum, attaching to the lower uterine segment. The round ligaments are attached to the anterior aspect of the uterus, originating just below the fallopian tubes. These ligaments support the uterus anterolaterally as they course into the internal inguinal ring to attach to the labia majora. The vesicouterine ligament extends from the cervix and attaches to the urinary bladder and has less supportive function for the uterus compared with the other three mentioned paired ligaments. Because most of the support is provided at the base, the remainder of the uterus is mobile and may change position between examinations and on occasion during an examination. The uterus is typically anteflexed with the fundus positioned anteriorly in relation to the cervix at an obtuse angle (see Figure 5-15 ). When the angulation of the cervix approaches 90 degrees, it is considered anteverted. The uterus is retroverted when the cervix is directed more vertically. When the body is flexed posteriorly in relation to the cervix, it is retroflexed ( Figure 5-21 ). Approximately 20% of women have a retroverted uterus. The uterus can also be lateroflexed, requiring two or three images to encompass it in the axial and sagittal planes. This is best demonstrated on coronal images ( Figure 5-22 ). Oblique imaging along the axis of the uterus may be obtained to better evaluate a uterus with variant position.

FIGURE 5-21 Uterine retroposition. T2-weighted sagittal images demonstrate varying degrees of posteroflexion of the uterus in two different patients. A, There is a large angle between the lower uterine segment and vagina with minimal posterior flexion at the body. B, Uterus is more retroverted with more significant posterior flexion at the body.

FIGURE 5-22 Lateral flexion of the uterus. Note the cervix (black arrowhead) is right of midline with the uterus body toward the left of midline (white arrowhead) . Note ovary just above the cervix. This variant anatomic relationship is well demonstrated on this coronal image.
The vascular supply of the uterus is through the paired uterine and ovarian arteries. The uterine arteries traverse the cardinal ligaments at the isthmus, providing branches anteriorly and posteriorly. Lymphatic drainage of the upper uterine corpus is provided by the lumbar paraaortic lymph node chain, whereas the lower body and cervical lymphatics drain into the parametrial, sacral, and iliac lymph nodes.

Imaging Appearance
The uterus during the reproductive age can be stratified into three zones best depicted on T2-weighted imaging, differentiated by the varying degrees of intracellular and extracellular fluid: the outer myometrium, inner myometrium (better known as the junctional zone), and the endometrium ( Figure 5-23 ). 5 - 7 The outer myometrium, also known as the stratum vasculare, has less compacted smooth muscle bundles with more interstitial spaces and is more vascular, demonstrating intermediate signal intensity on T2-weighted images. 7 This layer varies in thickness and appearance according to hormonal status and age, with the greatest growth during fertility and pregnancy, and regression after menopause ( Figure 5-24 , A and B ). 8 During the secretory phase and with the use of oral contraceptives, the outer myometrium appears brighter on T2-weighted images.

FIGURE 5-23 Zonal anatomy of the uterus. T2-weighted sagittal image best demonstrates the zonal anatomy with the central high signal region (E) representing the endometrium. This is immediately surrounded by a band low signal (J) that corresponds to the junctional zone or inner myometrium. This bands fades into the peripheral intermediate to high signal zone (O) composed of the outer myometrium. Areas of high signal correspond to slow-moving blood within veins found abundantly in this region. The outer thin line of low signal corresponds to the serosal lining.

FIGURE 5-24 A, Uterus of a 71-year-old woman. The junctional zone (star) becomes less well defined during menopause. Note relative atrophy of the uterine body relative to the uterine cervix. B, Uterus of a 94-year-old woman. Note that the uterine body ratio to the cervix decreases (arrows) . The junctional zone cannot be differentiated in this atrophic uterus. High signal structures within the myometrium are dilated slow-flowing veins.
The junctional zone represents the basal layer of the myometrium, which consists of more compacted longitudinally arranged smooth muscle bundles resulting in a denser nuclear ratio compared with the outer myometrium. 5 - 7 This results in low T2-weighted signal caused by relatively less vascularity and interstitial spaces compared with the outer myometrium. 9 The thickness of the junctional zone averages from 2 to 11 mm and is more apparent during the secretory phase and with the use of oral contraceptives because of increased brightness of the outer myometrium. It becomes slightly thicker and ill defined toward the end of the menstrual cycle 10 and becomes thinner with the use of oral contraceptives. 11 As with the rest of the myometrium, the junctional zone undergoes progressive atrophy during the postmenopausal period. With further lowering of T2 signal of the myometrium during this stage, the zonal anatomy becomes even more difficult to distinguish (see Figure 5-24 ). 8
On T2-weighted images the endometrium is bright in signal during all phases of the menstrual cycle with a variable thickness measuring up to 15 mm during the reproductive age. The glandular epithelium and endometrial fluid contribute to the high signal appearance on T2-weighted images best demonstrated in the sagittal plane ( Figure 5-25 ), with changes in thickness that correspond to the phase of the menstrual cycle. 10, 11 It measures approximately 5 to 10 mm during the secretory phase and is thickest just before menstruation. During menses a blood clot may be seen within the endometrial cavity. The endometrium is thinnest just after menses during the early proliferative phase. It can remain constantly thin in appearance (measuring up to 2 mm) with the use of birth control medications and during the postmenopausal period. The upper limit of endometrial thickness during the postmenopausal period is generally accepted as 5 mm, including those women taking exogenous estrogen or estrogen agonists.

FIGURE 5-25 A, Endometrial changes with menstrual cycle. Proliferative phase. T2-weighted sagittal image of the uterus demonstrates a thin endometrium. This is noted at the completion of menstruation and during early proliferative phase, which gradually becomes thicker as the cycle progresses. Note high signal in the deep myometrial layer (arrow) corresponding to slow blood flow in the veins. Small amount of fluid is also in the cul-de-sac (star) B, Secretory phase. T2-weighted sagittal image of the uterus demonstrates a thicker central high signal region corresponding to the endometrium. It can reexpand to 15 mm in thickness. Note cesarean section scar in the lower anterior wall (arrow) .
On T1-weighted images the endometrium and myometrium both demonstrate intermediate signal intensity, limiting differentiation between these layers. After administration of intravenous contrast ( Figure 5-26 ), the zonal anatomy of the uterus is better depicted with the outer myometrium demonstrating early enhancement, shortly followed by enhancement of the endometrium. This is in contradistinction to the junctional zone, which remains low in signal because of its more compact composition and relative sparse interstitial spaces. 6, 7

FIGURE 5-26 Dynamic imaging of the uterus. A, On T1-weighted imaging, myometrium and endometrium are isointense. Low signal in the endometrial cavity corresponds to high signal on T2-weighted image (E) and is related to fluid. The anterior intramural fibroid is isointense to the myometrium. B, T1-weighted early postcontrast image shows patchy enhancement of the fibroid and early patchy enhancement of the outer myometrium. C, Fibroid, outer myometrium, and endometrium demonstrate progressive enhancement. D, Delayed enhanced T1-weighted image. There is relative hypoenhancement of the junctional zone (black arrows) compared with the fibroid, outer myometrium, and endometrium (E) . The homogenously enhancing fibroid would be amenable to treatment with embolization. E, T2-weighted image of same patient shows high signal in endometrial region is related to higher fluid content. Low T2 signal of anterior intramural fibroid reflects dense collagen content.
Evaluation of the myometrium and endometrium is best performed with imaging in the sagittal plane. Not only does it optimally depict the zonal stratification of the myometrium and thickness of the endometrium, but this plane also demonstrates the relationship of the uterus with the rectum and urinary bladder, as well as the extent of disease.

Differential Considerations
Uterine fibroids are not uncommon, occurring in more than 40% of females above the age of 40. The usual appearance when small is a homogenously low T2 signal mass related to high collagen content. When larger, they become more heterogenous in signal as a result of degeneration. 12 They are typically intramural in location but can be submucosal or subserosal. Multiplanar T2-weighted imaging allows for more accurate localization of these masses, allowing for differentiation of the fibroids from normal myometrium and the endometrium. Before and after contrast T1-weighted imaging is the standard for uterine fibroid assessment to assess for the vascularity of the fibroids and their blood supply, especially when therapy is being considered ( Figure 5-27 ). Mimics include adjacent loops of bowel, fibrous adnexal tumors, focal adenomyosis, and dilated blood vessels.

FIGURE 5-27 Uterine leiomyoma. A, Sagittal T1-weighted precontrast image demonstrates a pedunculated subserosal fibroid (F) . Note that the endometrium is isointense with the remainder of the uterus (arrows) . Fibroid is slightly higher in signal but isointense with skeletal muscle. B, Sagittal T1-weighted early postcontrast image demonstrates more conspicuous endometrium as a result of relative increased enhancement of the myometrium (arrow) C, Sagittal T1-weighted late postcontrast image demonstrates lack of enhancement of the fibroid (F), indicating central degeneration. Endometrium is indicated by E.
Adenomyosis is a painful condition caused by extension of endometrial tissue into the myometrium with associated muscular hypertrophy. It is best depicted on multiplanar high-resolution T2-weighted images as focal or diffuse widening of the junctional zone (>11 mm) with indistinct margins, and often with tiny areas of high signal intensity resulting from ectopic endometrial glandular tissue ( Figure 5-28 ). Slow-moving blood in dilated veins should not be confused with these associated tiny cysts. Focal myometrial contractions that transiently decrease blood to an area of the uterus can mimic the appearance of adenomyosis. 13 Repeating the scan after a few minutes will reveal a return to normal uniform junctional zone thickness. Cinematic display using a half-acquisition RARE sequence shows the changes in the signal appearance of the myometrium and junctional zone related to focal contraction, but does not demonstrate an advantage over standard T2-weighted RARE imaging for most clinical work. 14 Recent work has shown that anticholinergics are effective in suppressing both uterine and intestinal contractions. 15

FIGURE 5-28 Adenomyosis. A, Diffuse adenomyosis. Sagittal T2-weighted image demonstrates diffuse thickening of the junctional zone (arrows) , which is indistinct and contains tiny cystic spaces secondary to ectopic endometrial tissue. Small intramural fibroid noted at the lower uterine segment anteriorly. B, Focal thickening of the junctional zone posteriorly (arrow) consistent with focal adenomyosis.
Richly vascularized endometrial polyps, endometrial cancer, and endometrial hyperplasia often resulting from exogenous hormones can cause abnormal thickening of the endometrium ( Figure 5-29 ). 16 Endometrial fluid in the form of blood, pus, or retained secretions can also make the endometrium prominent, mimicking other pathologies. 17 Large masses can cause the adjacent myometrium to become thin. 18, 19 Benign masses, such as fibroids and polyps, and anatomic changes associated with postmenopausal uterus, such as atrophy and loss of zonal distinction, can affect MRI assessment of endometrial and myometrial disease. 20 Familiarity with the normal appearance of the myometrium is essential when diagnosing and staging uterine malignancies.

FIGURE 5-29 Postmenopausal endometrial thickening. T2-weighted sagittal image shows irregular endometrial thickening (E) , with loss of sharpness at the lower junctional zone in a patient with endometrial carcinoma. Note indistinctness of the vaginal mucosa and paucity of periurethral vaginal plexus, which are not uncommon in this age group. Arrow, Urethra; V, vagina; R, rectum.


The cervix is referred to as “the neck of the uterus.” It is a cylindrical structure approximately 3 cm long and 2.5 cm in diameter, forming the narrowest part of the uterus that protrudes into the upper vagina. The cervical junction with the uterine body is termed the internal os , and the opening with the vaginal canal is the external os . It is mainly formed by a thick ring of dense collagenous connective tissue surrounded by circularly oriented smooth muscle that is continuous with the myometrium of the uterine corpus. Internal to this thick fibrous stromal ring is the endocervical canal, which is lined with mucus-secreting tall columnar epithelium arranged in folds known as the plicae palmatae, along with numerous mucous glands. The epithelial surface of the cervix that protrudes into the vagina (endocervix) is lined by stratified squamous epithelium. Although the epithelium of the endocervical canal does not vary with the menstrual cycle, it produces more viscous mucus during pregnancy and in the luteal phase of menstruation. The secretions become thinner during ovulation.
Support of the cervix is provided by ligamentous structures that form most of the pelvic floor. Anteriorly the pubocervical ligament and paired uterovesical ligaments provide support, whereas posteriorly the paired uterosacral ligaments support the cervix. The cardinal ligaments that anchor the cervix to the pelvic sidewalls provide most of the lateral support.

Imaging Considerations
The cervix is best demonstrated on axial T2-weighted imaging where it appears as a uniformly thick dark ring related to the sparse water content of its collagenous stroma ( Figures 5-30 and 5-31 ). This ring is continuous with the dark junctional zone layer of the uterine corpus. The fibrous stromal ring shows little variation with the menstrual cycle. Profound changes in the matrix composition and alignment of the collagenous fibers, however, lead to its increased elasticity during childbirth. Based on MRI studies, the cervical thickness starts to decrease, along with the uterine volume, after reaching maximum dimensions during the perimenopausal period. 21 This low signal ring surrounds a centrally located area of high signal related to the higher water content present in the cervical mucosa, mucus secretions, and invaginating crypts of the plicae palmatae. 22 Using a phased-array coil and high-resolution T2-weighted techniques discussed earlier, the plicae palmatae may be seen as corrugated intermediate signal bands arising from the center of the cervix ( Figure 5-32 ). These are more commonly seen between the third and fifth decades, 23 and can be mistaken for a cervical septum. 24 The outer circularly arranged layer of noncompact smooth muscle that is continuous with the outer myometrium of the uterine corpus is shown as a thin intermediate signal layer. 22, 25 Although non–contrast-enhanced T1-weighted images are not helpful in differentiating the layers of the cervix, strong enhancement of the cervical mucosa after contrast administration allows it to be separated from the lesser enhancing fibrous stroma.

FIGURE 5-30 Normal cervix zonal anatomy. T2-weighted sagittal image of the cervix shows circumferential low signal representing thick connective tissue cervical stroma (arrows) . This surrounds a faint high signal region of the endocervical mucosa. The outermost higher signal region is contiguous with the outer myometrium and is composed of more loosely organized connective tissue and muscle. Note that the thick cervical stroma is continuous with the junctional zone of the uterine body (J) .

FIGURE 5-31 Normal cervix. A, T2-weighted axial image shows the low signal endocervical stroma surrounding the high signal mucosa and its secretions. The less compact and more vascular outer cervix is intermediate in signal. Note that the perivaginal plexus is more intense in signal as a result of its slow-moving blood flow. B, T1-weighted postcontrast axial image demonstrates area of low signal corresponding to poorly enhancing cervical stroma, which surrounds low signal endocervical canal fluid. The remaining cervix, which is more vascular and less fibrous, and perivaginal plexus are both intermediate signal and are difficult to distinguish.

FIGURE 5-32 Coronal (A) and sagittal (B) T2-weighted images show the intermediate signal intensity plicae palmatae extending into the cervical canal (arrows) .

Differential Considerations
Some of the deep crypts of the endocervical mucous glands can become obstructed, leading to the formation of rounded inspissated mucus-filled areas of variable T2-weighted signal called nabothian cysts ( Figure 5-33 ). The squamocolumnar junction near the external os is an important area where most cervical carcinomas develop. On T2-weighted images, carcinomas are seen as an intermediate signal mass that disrupts the integrity of the low intensity fibrous stromal ring ( Figure 5-34 ). Assessment of the cervix can be limited by cervical angulation or displacement, and by the presence of nabothian cysts. Cervical assessment can be improved by decreasing slice thickness and obtaining images perpendicular to the long axis of the cervix. 26 In some cases the use of endovaginal gels can better depict the anatomy of the cervix and vagina. Although of limited sensitivity, MRI is highly specific for the detection of cervical involvement in endometrial disease. 27

FIGURE 5-33 Nabothian cysts. T2-weighted axial (A) and sagittal (B) images show more rounded and variable size areas of high signal throughout the endocervical canal compatible with nabothian cysts (arrow) . These may extend deeply into the cervical stroma.

FIGURE 5-34 Cervical carcinoma. T2-weighted sagittal image through the cervix shows an intermediate signal mass (arrows) disrupting the normally intense low signal ring of the cervical stroma with areas of high signal. The mass infiltrates the upper vagina (star) .

The Adnexa
The term adnexa is Latin for “the appendage of an organ.” These structures include the fallopian tubes, the ovaries, and their ligaments. The fallopian tubes functionally link the ovaries to the uterus. They extend from the cornual regions of the uterine corpus toward the ovaries, and are enveloped by the superior free border of the broad ligament. Each tube is approximately 10 to 12 cm long and is divided into four segments from the ovary to the uterus: infundibular, ampullary, isthmic, and interstitial segments. Adjacent to the ovary, the infundibular portion is dilated and contains frondlike projections facilitating transport of the ovum into the tube. The ampullary portion comprises the longest segment and is serpiginous in its course. The isthmus is a short segment that is narrowed as a result of progressive thickening of its muscular wall, whereas the interstitial or intramural portion traverses the uterine wall to open into the endometrial cavity.
The ovaries are paired almond-shaped organs, each measuring an average size of 3 × 1.5 × 1 cm and volume averaging 9.8 mL, with normal upper limit of 21.9 mL or approximately 4-cm diameter during the reproductive age. They are typically located in the ovarian fossae, an area between the internal and external iliac vessels, anchored by the broad ligament and round ligament posterolaterally and anteriorly, respectively. With each pregnancy or with uterine enlargement caused by fibroids, however, the ligaments may become lax, and the ovaries can be displaced more posteriorly or superiorly.
Each ovary is composed of a peripheral cortex and a central medulla. The medulla contains the majority of blood vessels, nerves, and lymphatics. The ovarian follicles reside in the cortex. All these structures are embedded in the ovarian stroma consisting of dense connective tissue. The ovaries undergo changes in size and appearance with age, menstrual cycle, and hormonal status. In response to hormonal stimulation of the menstrual cycle, a single dominant follicle develops from the group of follicles stimulated at the start of the cycle. As the other follicles undergo atresia, the dominant follicle reaches approximately 2.5 cm in size at midcycle and protrudes from the surface of the ovary. After ovulation the dominant follicle remnant involutes into the corpus luteum, which is incorporated back into the cortex. Hemorrhage into the corpus luteum cyst can occur owing to its rich vascularity. In the event that the dominant follicle fails to ovulate, a follicular cyst develops and can average up to 5 cm.
The ovarian artery supplies blood to the ovary and arises from the abdominal aorta at the level of the renal arteries. It anastomoses along the fallopian tube with the uterine artery. On the right the ovarian vein drains into the inferior vena cava, and on the left it drains into the left renal vein. These anatomic landmarks and relationships are important to aid in the identification of the ovaries, especially with atrophy of postmenopausal ovaries. The lymphatic drainage of the ovaries occurs via ovarian vessels to the lumbar chain of lymph nodes along the paraaortic region.

Imaging Appearance
The fallopian tubes are filled with fluid and contain many mucosal folds; however, because of their narrow width (1 to 2 mm in most segments), they are not identified on MRI unless pathologically dilated ( Figure 5-35 ). When the tubes are dilated as a result of obstruction or endometriosis, high signal filling on T2-weighted images can be seen 28 and is best confirmed with multiplanar imaging.

FIGURE 5-35 Hydrosalpinx. T1-weighted (A) and T2-weighted (B) coronal images demonstrate fluid within a dilated tube in the left adnexa (arrow) consistent with hydrosalpinx. It is important to identify the ovary as a separate structure to avoid the pitfall of mischaracterization as an ovarian cyst. C and D, Hydrosalpinx in a second patient showing more elongated appearance of the fallopian tube (arrow). Fibroid is marked with a star.
The ovaries can easily be seen in the vast majority of premenopausal women on T2-weighted images by identification of the high signal follicles located with low signal cortical stroma. On non–contrast-enhanced T1-weighted images, the ovaries demonstrate homogenously intermediate to low signal. With contrast administration the follicles are low on T1-weighted images and demonstrate thin variably enhancing walls in the background of enhancing stroma of the cortex and medulla ( Figure 5-36 ). During the reproductive ages, the medulla has a slightly higher signal compared with the stroma of the cortex on T2-weighted images, and then becomes more uniformly low in signal after menopause, reflecting the more compact connective tissue that occurs with the accumulation of remnants of ovulation. 29

FIGURE 5-36 Normal ovaries. A, T2-weighted axial image of the same patient as in Figure 5-11 demonstrates increased contrast between the follicles, which are hyperintense compared with low signal stroma. A mixed low-intermediate signal area in the left ovary may represent an evolving hemorrhagic cyst. B, T1-weighted precontrast axial image. Both ovaries demonstrate ill-defined follicles, which are slightly low in signal compared with ovarian stroma. C, T1-weighted immediate postcontrast. Note enhancement in the iliac arteries (arrows) D, T1-weighted postcontrast (3 minutes) axial image demonstrates increased contrast between the dark follicles and enhancing ovarian stroma (white arrows) . Note pelvic varices (curved arrow) . The ovarian stroma enhances less than the myometrium.
Hormonal stimulation from the normal menstrual cycle or from external sources influences the appearance of the ovaries. The peripherally located follicles are smaller at the beginning of the cycle when the ovaries are smaller. The ovaries enlarge as the follicles begin to develop with a dominant follicle becoming larger than the rest. When the dominant follicle fails to ovulate, a follicular cyst develops, demonstrating thin rim enhancement. Corpus luteal cysts typically have regularly thick walls that enhance and may be crenated in appearance after rupture ( Figure 5-37 ), demonstrating high signal on both T2- and T1-weighted images compatible with hemorrhage. 30 Follicles reach maximum numbers during the third and fourth decades. The mean volume of follicles decreases after the fifth decade, 31 becoming more homogenous in appearance, with low to intermediate signal on both T1- and T2-weighted images. Because of the lack of hormonal stimulation, however, preexisting follicles can persist several years after menopause and persist as high signal cysts on T2-weighted images.

FIGURE 5-37 Corpus luteal cyst. T2-weighted axial images without and with fat saturation (A and B) demonstrate a thick-walled cystic mass in the left ovary (arrows) C and D, Before and after contrast T1-weighted axial images demonstrating enhancement of the thick wall of the cyst. The wall is gently undulating, a finding seen with corpus luteal cysts undergoing resorption (arrows) .

Differential Considerations
Occasionally congested pelvic veins will mimic an ovary, owing to high T2 signal of slow-flowing blood. Lack of definable cortex in an orthogonal imaging plane excludes the presence of an ovary. Although paraovarian cysts demonstrate similar signal intensity characteristics as ovarian cysts, preservation of the normal ovarian contour when they are adjacent to each other helps to differentiate the two. 32 Peritoneal inclusion cysts have a tethered appearance to their margins and may distort the configuration of the ovary adjacent to it. 33 Endometriomas involving the ovary have unique signal characteristics that allow them to be distinguished from solid masses (see Figure 5-7 ). In contrast to hemorrhagic cysts, they tend to be multifocal, and because they contain blood components of varying ages, they are more heterogenous in signal. Fat, as seen with dermoids, is also bright on T1-weighted images and could be distinguished from hemorrhagic cysts using fat saturation techniques. With uniformly low signal on T2-weighted imaging, a small pedunculated fibroid may mimic the appearance of an ovarian fibroma. Features seen in adnexal cysts associated with malignancies include larger size, soft tissue nodularity, thickened septations, and calcifications.

Magnetic Resonance Imaging Appearance of Adjacent Pelvic Structures

Urinary Bladder and Urethra
The urinary bladder and the urethra are contained within the anterior compartment of the pelvic cavity, located adjacent to most of the length of the anterior vaginal wall with the exception of the uppermost portion formed by the cervix. The wall of the urinary bladder is isointense, with skeletal muscle rendering it hypointense on T2-weighted images and hypointense to isointense on T1-weighted images. 34 The bladder mucosa demonstrates enhancement on postcontrast T1-weighted images, and is slightly more intense in signal to the muscular wall on T2-weighted images. Although the dome of the bladder is partly surrounded by fat and draped by overlying peritoneum, the base of the bladder is separated from the vagina by a richly vascular connective tissue, which is hyperintense on T2-weighted images. The dome and anterior and posterior walls of the bladder are best evaluated on sagittal and parasagittal images, whereas the lateral walls are better depicted in the coronal plane. The degree of urinary bladder distension may limit the assessment for wall irregularities and can influence the orientation of the uterus. Because of fat surrounding most of the urinary bladder, chemical shift artifacts can occur, creating the appearance of thickening of the wall on one side and thinning of the wall on the contralateral side. This artifact can be minimized by changing the direction of the frequency-encoding gradient or with the use of fat suppression techniques. Concentrated gadolinium chelate appears distinctly hypointense to urine on T2-weighted images, and because of its higher density is seen more dependently in the urinary bladder. Because it is in the irradiated field in pelvic malignancies, urinary bladder wall thickening and enhancement may be observed posttreatment and should be recognized ( Figure 5-38 ).

FIGURE 5-38 Postradiation cystitis. T2-weighted (A) and T1-weighted (B) postcontrast sagittal images demonstrate diffuse urinary bladder wall thickening and with mucosal enhancement in a patient with pelvic radiation treatment. Note the changes of the bone marrow signal in the lumbosacral spine consistent with marrow edema, and presacral thickening.
The female urethra is located anterior to the lower vagina and measures approximately 4 cm. The walls of the urethra are best depicted with T2-weighted images in the axial plane demonstrating an intermediate-signal round structure corresponding to the external sphincter, with central punctuate hyperintensity from the redundant mucosal epithelium. 34, 35 The submucosal layer is higher in signal from slow flow of its vascular network interposed in the areolar connective tissue called the perivaginal fascia . 36 The surrounding external sphincter composed of striated muscle is hypointense in signal. These structures are difficult to discern on T1-weighted images but are better seen with contrast enhancement with intense submucosa enhancement surrounding a central low-signal dot.
The volume of the paravaginal fascia is associated with urinary incontinence, more prevalent in individuals with reduced volume. 37 Obstructed periurethral glands may rupture into the urethra, resulting in the formation of diverticula, which may assume lobular or circumferential configurations. These more commonly involve the posterior lateral mid aspect of the urethra 38 and are best characterized using multiplanar high-resolution MRI 39 (see Figure 5-18 ) with postcontrast T1-weighted images.

Pelvic Bowel
Located much further from the diaphragm, pelvic loops of bowel are less affected by respiratory motion artifacts. With the advent of faster imaging sequences, motion artifacts resulting from intestinal peristalsis are minimized. When this becomes problematic, however, reversal of the phase-encoding direction can be used to minimize artifacts related to peristalsis. Normally the walls of the small and large bowel and rectum demonstrate intermediate signal on T1-weighted images and low signal on T2-weighted images. The intraluminal contents demonstrate variable signal intensity related to content with signal void from air and high T2 signal with water. Both positive (oil or gadolinium) and negative (air) contrast agents may be used to evaluate mucosal abnormalities. Postcontrast T1-weighted images depict mucosal and submucosal enhancement, which are inseparable, demonstrating a thin enhancing wall. This sequence, when combined with fat saturation techniques and multiplanar acquisition, is an important advantage in detecting intestinal pathology. On the other hand, the anal canal, which is composed of a thicker muscular wall from the external sphincter, demonstrates the zonal anatomy with better detail. The rectum is separated from the posterior vaginal wall by a thin layer of richly vascularized connective tissue and fat, which is high in signal on T2-weighted images. The anterior aspect of the mid to upper rectum is lined by a peritoneal covering as it forms the posterior border of the pouch of Douglas. Similar to the urinary bladder, wall thickening from postradiation changes can be seen involving pelvic loops of bowel.
Tiny amounts of peritoneal fluid tend to collect in the cul-de-sac and are best evaluated in the sagittal plane. A small amount of free fluid in the cul-de-sac is likely physiologic in an asymptomatic female patient, related to the menstrual cycle. When a large amount of fluid is present, concern for metastasis is raised in the appropriate clinical setting. Enhancement of the peritoneum is nonspecific and most often related to inflammatory states or malignancy. Irregularity of the cul-de-sac and the anterior rectal surface may be seen with peritoneal spread of metastatic disease or in cases of endometriosis ( Figure 5-39 ) is often better depicted with fat suppression techniques. Sampling of this fluid can be performed with a culdocentesis.

FIGURE 5-39 T2-weighted sagittal image demonstrates large amount of ascites (A) filling the pelvis and compressing the urinary bladder (B) . Note the serosal surface of the uterus. Normally the cul-de-sac dips more caudally approaching the posterior vaginal fornix. In this case of known peritoneal metastatic disease, however, there is filling of the cul-de-sac with low signal consistent with a metastatic focus, which tends to aggregate in this most dependent portion (curved arrows) .

Osseous Structures
Cortical bone demonstrates dark signal on both T1- and T2-weighted images and should be uninterrupted in contour. The appearance of bone marrow is more variable and is related to age, specifically more heterogenous in older populations. Focal areas of high T1-weighted signal involving vertebral bodies may be secondary to fat islands caused by focal conversion or hemangiomas with high fat content, and can be confirmed by using fat suppression techniques. Other processes that affect bone marrow appearance are anemia, chemotherapy, radiation, and metastatic disease. Hematopoietic marrow appears hypointense on T1-weighted images, and becomes progressively hyperintense with advancing age from progressive marrow conversion to fat. In patients with anemia, marrow fat is reconverted to red marrow, demonstrating diffuse loss of signal on T1-weighted images without corresponding high signal on T2-weighted images. Conversely, metastatic foci are usually low in signal on T1-weighted images and moderately high in signal on T2-weighted images. Generally disease in the spine may be seen as lower signal compared with the intervertebral disks, or can be isointense to muscle. Response to chemotherapy includes reactivation of marrow with conversion to low signal on T1-weighted images as replaced by hematopoietic cells. Postradiation changes to the spine range from marrow edema early after treatment to fatty replacement later in the course of therapy. It is best depicted in the sagittal plane as a sharp margination of increased T1-weighted signal of the marrow involved in the treatment port (see Figure 5-38 ). Metal from spine and hip orthopedic hardware, as well as surgical clips, can cause artifactual signal void in the field with distortion obscuring regional anatomy.

Pelvic Floor Muscles
The female pelvic cavity can be subdivided into three functional compartments. The urinary bladder and urethra comprise the anterior compartment; the uterus, vagina, and ovaries make up the middle compartment; and the rectum occupies the posterior compartment. A framework of muscle, fascia, and ligaments provides support for these compartments. Integral to the pelvic floor is the perineal body, which is a fibromuscular mass situated in the middle of the perineum, between the vagina and anus. The perineal body serves as an attachment point for many pelvic floor muscles. The support for the posterior compartment is mainly provided by the levator ani, a thin broad muscle that is composed of the iliococcygeus, pubococcygeus, and puborectalis muscles. The iliococcygeus is a horizontal broad muscle that originates from the tendinous arch of the pelvic sidewall inserting posteriorly in the midline at the anococcygeal raphe, forming the levator plate. The pubococcygeus extends from the pubis to the coccyx and supports the pelvic floor like a hammock. The puborectalis provides support mainly for the middle and anterior compartments. This muscle arises from the inferior aspect of the symphysis pubis and forms a strong sling around the lower part of the rectum, providing support for the vagina and bladder neck. In the resting contracted state of the pelvic floor muscles, the pelvic organs are supported and compressed anteriorly, preventing visceral prolapse and contributing to urinary continence.
The visceral and pelvic fascia, along with fascial condensations known as ligaments, also provides support for the pelvic floor and organs. The endopelvic fascia is a thin delicate layer that extends from the abdomen, draping over the pelvic floor. In the middle compartment, the uterus and vagina receive support from the endopelvic fascia, uterosacral ligament, and paracolpium, preventing prolapse. Anteriorly the urinary bladder and uterine cervix are partly supported by the pubocervical fascia, preventing cystoceles and urinary incontinence. In the posterior compartment, the rectum receives support from the rectovaginal fascia and posterior vaginal wall, which when compromised may lead to an enterocele or rectocele.
Larger muscles surrounding the pelvic cavity, such as the iliopsoas, internal obturator, and pyriformis muscles, are clearly seen as low signal structures on T1- and T2-weighted images. These muscles may demonstrate abnormal T2-weighted high signal with inflammation or tumor. T1-weighted sequences are helpful for the evaluation of fatty replacement resulting from paralysis ( Figure 5-40 ), and hemorrhage.

FIGURE 5-40 Pyriformis syndrome. A, T2-weighted rapid acquisition with relaxation enhancement (RARE) axial image showing increased signal infiltrating the pyriformis muscles bilaterally and, to a lesser extent, the neighboring gluteal muscles in this patient with complaints of buttock pain (arrows) . Findings are compatible with pyriformis syndrome from inflammation of the belly of the pyriformis causing local inflammation and compression of the sciatic nerve as it passes underneath the muscle belly. Patient had a history of radiation treatment to the pelvis. B, T2-weighted inversion recovery axial image with suppression of fat shows edema and inflammation more dramatically compared with standard RARE shown in A (arrows).
When evaluating the pelvic floor as part of the workup for incontinence, most of the fascial condensations of the pelvic floor are not directly visualized on MRI; however, the levator ani is clearly identified on rapid T2-weighted images. 40 Assessment of the urogenital and pelvic diaphragm is enhanced with use of an endovaginal coil. 41 Pelvic floor integrity can be assessed by obtaining multiple rapid midline sagittal images at rest and during Valsalva maneuver, which will demonstrate hypermobility of the pelvic organs.

Lymph Nodes
Normal retroperitoneal and pelvic lymph node size is usually 3 to 5 mm in the short axis. 42 Any lymph node with a short axis dimension more than 10 mm is suspicious for metastatic involvement. Normal nodes are typically elongated in appearance and may contain a fatty hilum, whereas pathologic lymph nodes are usually spherical in configuration. When visualized, normal lymph nodes are typically low in signal on T1-weighted images and intermediate to slightly higher in signal on T2-weighted images in the background of fat. When pathologically enlarged, they are best detected with post–contrast-enhanced T1-weighted images, where they are usually brighter than adjacent organs. T2-weighted images are also helpful to detect necrotic lymph nodes that demonstrate high signal centrally. Multiplanar views should be carefully reviewed when assessing for lymph node enlargement. Obturator nodes are often best seen in the lateral images of the sagittal series. Slightly hyperintense ring flow artifact found in dilated iliac veins with slow flow and loops of bowel in femoral and inguinal hernias may be confused with adenopathy. Because both reactive and metastatic lymph nodes can enhance similarly after administration of gadolinium-based intravenous contrast agents, magnetic resonance lymphography, which uses ultrasmall particles of iron oxide, may show promise for the detection of lymph node metastasis with high specificity. 43 More recent work, however, shows limitations in detecting metastasis in smaller lymph nodes or micrometastasis. 44

As described in separate sections earlier, the pelvic viscus receives most of its blood supply from the internal iliac artery, which branches off the common iliac artery anterior to the sacroiliac joint. It further subdivides into anterior and posterior divisions, with the anterior division providing blood supply to most of the pelvic viscera. Anterior branches are the uterine, vaginal, inferior vesical, obturator, middle rectal, internal pudendal, and inferior gluteal arteries. The posterior division gives rise to the iliolumbar, lateral sacral, and superior gluteal arteries.
In the evaluation of uterine masses, source data obtained from 3D contrast images can be used to reconstruct the vascular anatomy that would be useful for surgical and endovascular treatment planning (see Figure 5-12 ). It is important that the vessels evaluated are within the imaging volume, and that axial source images are reviewed to minimize overestimation or underestimation of abnormalities seen on postprocessed maximum intensity projection images. When vascular structures are evaluated using nonintravenous contrast techniques (spin and gradient echo), it is important to recognize potential mimics of intraluminal thrombus, such as slow flow, in-plane flow, entry phenomenon, and thrombus in the acute or subacute stages.


Optimal MRI of the pelvis requires magnet strength of 1.5 Tesla or greater, proper positioning of a body array coil, and appropriate choice of pulse sequence parameters, particularly field of view.
T2-weighted images in the sagittal and coronal planes yield the most valuable information and should be performed early in the examination.
Identification of the uterus, cervix, and ovaries is possible in virtually every woman of menstrual age because of the inherent contrast properties of these organs.
Contrast-enhanced imaging using T1-weighted 2D or 3D gradient echo is not necessary for accurate depiction of anatomic abnormalities.


1. Troiano R.N., Lange R.C., McCarthy S. Conspicuity of normal and pathologic female pelvic anatomy: comparison of gadolinium-enhanced T1-weighted images and fast spin echo T2-weighted images. J Comput Assist Tomogr . 1996;20(6):871-877.
2. Lauenstein T.C., Sharma P., Hughes T., et al. Evaluation of optimized inversion-recovery fat-suppression techniques for T2-weighted abdominal MR imaging. J Magn Reson Imaging . 2008;27(6):1448-1454.
3. Hricak H., Chang Y.C., Thurnher S. Vagina: evaluation with MR imaging. Part I. Normal anatomy and congenital anomalies. Radiology . 1988;169(1):169-174.
4. Seigelman E.S., Outwater E.K., Banner M.P., et al. High-resolution MR imaging of the vagina. Radiographics . 1997;17(5):1183-1203.
5. Lee J.K., Gersell D.J., Balfe D.M., et al. The uterus: in vitro MR-anatomic correlation of normal and abnormal specimens. Radiology . 1985;157(1):175-179.
6. Scoutt L.M., Flynn S.D., Luthringer D.J., et al. Junctional zone of the uterus: correlation of MR imaging and histologic examination of hysterectomy specimens. Radiology . 1991;179(2):403-407.
7. Brown H.K., Stoll B.S., Nicosia S.V., et al. Uterine junctional zone: correlation between histologic findings and MRI imaging. Radiology . 1991;179(2):409-413.
8. Demas B.E., Hricak H., Jaffe R.B. Uterine MR imaging: effects of hormonal stimulation. Radiology . 1986;159:123-126.
9. McCarthy S., Scott G., Majumdar S., et al. Uterine junctional zone: MR study of water content and relaxation properties. Radiology . 1989;171(1):241-243.
10. Haynor D.R., Mack L.A., Soules M.R., et al. Changing appearance of the normal uterus during the menstrual cycle: MR studies. Radiology . 1986;161(2):459-462.
11. McCarthy S., Tauber C., Gore J. Female pelvic anatomy: MR assessment of variations during the menstrual cycle and with use of oral contraceptives. Radiology . 1986;160(1):119-123.
12. Hricak H., Tscholakoff D., Heinrichs L., et al. Uterine leiomyomas: correlation of MR, histopathologic findings, and symptoms. Radiology . 1986;158(2):385-391.
13. Togashi K., Kawakami S., Kimura I., et al. Uterine contractions: possible diagnostic pitfall at MR imaging. J Magn Reson Imaging . 1993;3(6):889-893.
14. Masui T., Katayama M., Kobayashi S., et al. Pseudolesions related to uterine contraction: characterization with multiphase-multisection T2-weighted MR imaging. Radiology . 2003;227(2):345-352.
15. Nakai A., Togashi K., Kosaka K., et al. Do anticholinergic agents suppress uterine peristalsis and sporadic myometrial contractions at cine MR imaging? Radiology . 2008;246(2):489-496.
16. Ascher S.M., Johnson J.C., Barnes W.A., et al. MR imaging appearance of the uterus in postmenopausal women receiving tamoxifen therapy for breast cancer: histopathologic correlation. Radiology . 1996;200(1):105-110.
17. Takeuchi M., Matsuzaki K., Uehara H., et al. Pathologies of the uterine endometrial cavity: usual and unusual manifestations and pitfalls on magnetic resonance imaging. Eur Radiol . 2005;15(11):2244-2255.
18. Sironi S., Colombo E., Villa G., et al. Myometrial invasion by endometrial carcinoma: assessment with plain and gadolinium-enhanced MR imaging. Radiology . 1992;185(1):207-212.
19. Yamashita Y., Harada M., Sawada T., et al. Normal uterus and FIGO stage I endometrial carcinoma: dynamic gadolinium-enhanced MR imaging. Radiology . 1993;186(2):495-501.
20. Scoutt L.M., McCarthy S.M., Flynn S.D., et al. Clinical stage I endometrial carcinoma: pitfalls in preoperative assessment with MR imaging. Work in progress. Radiology . 1995;194(2):567-572.
21. Hauth E.A., Jaeger H.J., Libera H., et al. MR imaging of the uterus and cervix in healthy women: determination of normal values. Eur Radiol . 2007;17(3):734-742.
22. Scoutt L.M., McCauley T.R., Flynn S.D., et al. Zonal anatomy of the cervix: correlation of MR imaging and histologic examination of hysterectomy specimens. Radiology . 1993;186(1):159-162.
23. Takahata A., Koyama T., Kido A., et al. The frequency of the plicae palmatae in the uterine cervix on MR imaging. Abdom Imaging . 2009;34(2):277-279.
24. El Jack A.K., Siegelman E.S. Pseudoseptum” of the uterine cervix on MRI. J Magn Reson Imaging . 2007;26(4):963-965.
25. DeSouza N.M., Hawley I.C., Schwieso J.E., et al. The uterine cervix on in vitro and in vivo MR images: a study of zonal anatomy and vascularity using an enveloping cervical coil. AJR Am J Roentgen . 1994;163(3):607-612.
26. Kinkel K. Pitfalls in staging uterine neoplasm with imaging: a review. Abdom Imaging . 2006;31(2):164-173.
27. Sanjuán A., Escaramís G., Ayuso J.R., et al. Role of magnetic resonance imaging and cause of pitfalls in detecting myometrial invasion and cervical involvement in endometrial cancer. Arch Gynecol Obstet . 2008;278(6):535-539.
28. Outwater E.K., Siegelman E.S., Chiowanich P., et al. Dilated fallopian tubes: MR imaging characteristics. Radiology . 1998;208(2):463-469.
29. Outwater E.K., Talerman A., Dunton C. Normal adnexa uteri specimens: anatomic basis of MR imaging features. Radiology . 1996;201(3):751-755.
30. Outwater E.K., Mitchell D.G. Normal ovaries and functional cysts: MR appearance. Radiology . 1996;198(2):397-402.
31. Hauth E.A., Jaeger H.J., Libera H., et al. Magnetic resonance imaging of the ovaries of healthy women: determination of normal values. Acta Radiol . 2006;47(9):986-992.
32. Kishimoto K., Ito K., Awaya H., et al. Paraovarian cyst: MR imaging features. Abdom Imaging . 2002;27(6):685-689.
33. Hoffer F.A., Kozakewich H., Colodny A., et al. Peritoneal inclusion cysts: ovarian fluid in peritoneal adhesions. Radiology . 1988;169(1):189-191.
34. Fisher M.R., Hricak H., Crooks L.E. Urinary bladder MR imaging. Part I. Normal and benign conditions. Radiology . 1985;157(2):467-470.
35. Hricak H., Secaf E., Buckley D.W., et al. Female urethra: MR imaging. Radiology . 1991;178(2):527-535.
36. Strohbehn K., Quint L.E., Prince M.R., et al. Magnetic resonance imaging anatomy of the female urethra: a direct histologic comparison. Obstet Gynecol . 1996;88(5):750-756.
37. DeSouza N.M., Daniels O.J., Williams A.D., et al. Female urinary genuine stress incontinence: anatomic considerations at MR imaging of the paravaginal fascia and urethra initial observations. Radiology . 2002;225(2):433-439.
38. Ryu J., Kim B. MR imaging of the male and female urethra. Radiographics . 2001;21(5):1169-1185.
39. Blander D.S., Rovner E.S., Schnall M.D., et al. Endoluminal magnetic resonance imaging in the evaluation of urethral diverticula in women. Urology . 2001;57(4):660-665.
40. Kim J.K., Kim Y.J., Choo M.S., et al. The urethra and its supporting structures in women with stress urinary incontinence: MR imaging using an endovaginal coil. AJR Am J Roentgenol . 2003;180(4):1037-1044.
41. Tan I.L., Stoker J., Zwamborn A.W., et al. Female pelvic floor: endovaginal MR imaging of normal anatomy. Radiology . 1998;206(3):777-783.
42. Grubnic S., Vinnicombe S.J., Norman A.R., et al. MR evaluation of normal retroperitoneal and pelvic lymph nodes. Clin Radiol . 2002;57(3):193-200. discussion 201–204
43. Barentsz Jo, Fütterer J.J., Takahashi S. Use of ultra small superparamagnetic iron oxide in lymph node MR imaging in prostate cancer patients. Eur J Radiol . 2007;63(3):369-372.
44. Keller T.M., Michel S.C., Fröhlich J., et al. USPIO-enhanced MRI for preoperative staging of gynecological pelvic tumors: preliminary results. Eur Radiol . 2004;14(6):937-944.

Suggested Readings

Fielding J.R. MR imaging of the female pelvis. Radiol Clin North Am . 2003;41(1):179-192.
Outwater E.K., Talerman A., Dunton C. Normal adnexa uteri specimens: anatomic basis of MR imaging features. Radiology . 1996;201(3):751-755.
Chapter 6 Diffusion Magnetic Resonance Imaging

Jurgen J. Fütterer
The value of magnetic resonance imaging (MRI) for evaluating a variety of gynecologic diseases, owing to its excellent tissue contrast-reflecting pathology, has been established. 1 Diffusion-weighted (DW) imaging allows the observation of differences in molecular diffusion caused by the random and microscopic motion of molecules (“brownian motion”). 2 Recent technical advances in DW imaging greatly enhanced the clinical value of MRI of the body. DW imaging plays an important role in the diagnosis of brain disorders, more specifically diagnosis of acute stroke. It has not been fully applied to body MRI because the images become more distorted by its sensitivity to motion, resulting in misregistration attributable to chemical-shift artifacts. 3 DW imaging can demonstrate abnormal signals emitted by pathologic foci based on differences in molecular diffusion. DW imaging can delineate malignant lesions displaying hyperintense signal because water diffusion is restricted in tissues of high cellularity. 4, 5 It also permits the quantitative evaluation of the apparent diffusion coefficient (ADC) that may be helpful for monitoring therapeutic outcomes. Decreased ADC values of malignant tumors compared with those of normal tissues or benign lesions have been previously reported. DW imaging has a potential role in distinguishing cancerous from normal tissues.

Description of Technical Requirements
DW imaging is obtained by measuring signal loss after a series of two motion-providing gradient pulses are added to both sides of a 180-degree refocusing radiofrequency (RF) pulse to enhance differences in molecular diffusion between tissues. 3 DW imaging estimates the mean distance traveled by all hydrogen nuclei in every voxel of imaged tissue. The greater this mean distance, the more self-diffusion of water molecules has occurred in a certain time interval. 6 The most common method used in clinical MRI is bipolar-gradient diffusion preparation with single-shot echo-planar imaging. 7, 8 The latter approach can be limited by susceptibility artifacts but can capture the diffusion contrast while minimizing both eddy current and motion artifacts. Recent developments in parallel imaging techniques have improved the quality of body DW imaging by reducing the acquisition time and by minimizing the echo-planar imaging-related susceptibility artifacts. 9, 10


In the female pelvis, DW imaging can be applied for tumor detection to differentiate between benign and malignant lesions and for the detection of lymph node and bone metastases. DW imaging may also have a potential role in monitoring therapeutic outcome.

There are no contraindications for a DW imaging sequence in the female pelvis.

Technique Description
DW imaging provides quantification of brownian motion of water protons by calculating the ADC, and can be used for in vivo quantification of the combined effects of capillary perfusion and diffusion. 2 The degree of restriction to water diffusion in biologic tissue is inversely correlated to tissue cellularity and the integrity of cell membranes. Free motion of water molecules is more restricted in tissues with a high cellular density. DW imaging with single-shot echo-planar imaging can provide excellent contrast-to-noise ratio because the signals of most organs are low, whereas the lesion signals remain high. 11, 12 The sensitivity of the DW imaging sequence to water motion can be varied by changing the gradient amplitude, expressed as the b -value. By performing DW imaging using different b -values, quantitative analyses can be performed to determine the apparent diffusion coefficient. Because diffusion in tissue is limited by cellular structures, to establish a reliable estimate of the mean distance traveled by the hydrogen nuclei, DW imaging is acquired in at least three orthogonal directions for each b -value. 6, 13 This phenomenon of varying restriction of self-diffusion along different axes is called anisotropy . As in linear aligned tissue, this anisotropy is more pronounced because there is one direction that contributes most to the diffusion. Diffusion tensor imaging is a technique that quantifies the level of anisotropy in tissue, expressed in a fractional anisotropy value. The latter can be used in addition to DW imaging to determine the structural organization of tissue along which diffusion takes place.

Pitfalls and Solutions
DW imaging typically has T2-weighted and DW characteristics. The intensity of the signal on the DW image represents a combination of signal from the T2 relaxation and the dephasing caused by the water motion in the presence of diffusion gradients. At low b -values (<100 s/mm 2 ) there is greater contribution from the T2 signal, and at high b -values contrast is determined by relative diffusion. 6 When a diffusion image is bright because of high T2 signal rather than restricted diffusion, it is known as T2 shine-through effect. On DW imaging with an intermediate b -value (e.g., 500 s/mm 2 ), urine or ascites appears as high-signal comparable to tumor. ADC maps should be obtained with at least two b -values, typically a low b -value (50 to 100 s/mm 2 ) and a high b -value (>500 s/mm 2 ). Tissue microperfusion can contaminate the signal attenuation in DW imaging acquisition, which can be decreased by choosing a low b -value greater than 0 s/mm 2 (preferably ≥50 s/mm 2 ).
To minimize the influence of bulk motion as a distorting factor and minimizing T2 shine-through, typically an echo time as short as possible, a smaller number of echo train lengths, and application of a parallel imaging technique are chosen. Although a wider receiver bandwidth reduces the signal-to-noise ratio, it is recommended because it shortens the duration of acquisition of the MRI signal and reduces susceptibility artifacts.
Ghosting from respiratory motion and chemical shift artifact can be reduced by fat suppression. The method of fat saturation can significantly influence the signal-to-noise ratio of the image and repetition time. In body regions, short time inversion recovery (STIR) fat suppression technique may provide more homogeneous fat suppression compared with chemical shift selective fat suppression. Consequently the signal-to-noise ratio STIR fat suppression technique may be prolonged because of the longer repetition time and an increased number of excitations, which is required for decreased signal-to-noise ratio.

Image Interpretation
There are two methods for assessment of DW imaging: (1) visual assessment of the DW images and ADC maps; and (2) quantitative assessment of ADC values, which gives more specific information regarding molecular diffusion. The ADC value can be derived on a voxel-by-voxel basis and depicted on an ADC map, which allows one to measure the ADC values of a specific region of interest.
Malignant tumors are generally depicted as foci of increased intensity on DW imaging because water diffusion is restricted in highly cellular tissue in malignant tumors. A variety of benign tumors, however, can show restricted diffusion, which may limit the role of DW imaging. Hence the ADC maps and DW images should be read in conjunction with the conventional MRIs. On DW imaging with a high b -value (i.e., >1000 s/mm 2 ), malignant tumors and lymph nodes are more conspicuous compared with images with lower b -values because most of the normal pelvic tissue is strongly suppressed.

Uterine Cervix
Normal cervical stroma contains large amounts of fibrous tissue, so that it demonstrates low signal intensity on T2-weighted MRIs and delayed enhancement on postgadolinium T1-weighted MRIs. Because of the greater cellularity of cervical cancer compared with the normal fibroelastic cervical stroma, tumor demonstrates higher signal intensity on T2-weighted images and greater enhancement on postgadolinium T1-weighted MRIs. The ADC of cervical cancers is significantly lower than the median ADC of normal cervix stroma, with little overlap. Cervical cancer has been shown to demonstrate impeded diffusion relative to normal cervical stroma. Examples of cervical cancer images using diffusion imaging are seen in Figures 6-1 through 6-3 . Distinction between histologic tumor grades and subtypes based on ADC is not possible. This may be the result of factors such as nuclear size and necrosis. ADC may have predictive value in squamous tumors. Furthermore, the mean ADC value of cervical cancer has been reported to be significantly increased after chemotherapy and/or radiation therapy. Further study into its predictive value for long-term outcome will determine the ultimate clinical value.

FIGURE 6-1 Stage IIB cervical cancer. Axial T2-weighted turbo spin echo (A) and sagittal fat-suppressed T1-weighted postcontrast (B) magnetic resonance images (MRIs) in a young patient with cervical carcinoma demonstrate a mass with parametrial invasion on the right side (arrows) . Rectum marked with star. The tumor is easily appreciated as a mass of high signal intensity on axial diffusion MRI ( b = 800) (arrows) (C) with a corresponding area of restricted diffusion on the apparent diffusion coefficient map (arrows) (D) .

FIGURE 6-2 Ovarian cancer recurrence. Axial T2-weighted turbo spin echo magnetic resonance image (A) in a patient with increasing cancer antigen-125 level shows a small intermediate signal intensity lesion in the meso-rectosigmoid fat (arrow) . The deposit is better appreciated on the diffusion-weighted ( b = 800) (arrows) (B) and apparent diffusion coefficient (C) images (star) .

FIGURE 6-3 Stage IIB cervical cancer. Axial T2-weighted image (A) magnetic resonance image (MRI) in a patient with cervical carcinoma demonstrates a mass with parametrial invasion on the ventral side (arrow) . A small fat line is visible between the tumor and the bladder wall. The tumor is easily appreciated as a mass of high signal intensity on axial diffusion MRI ( b = 800) (arrow) (C) with a corresponding area of restricted diffusion on the apparent diffusion coefficient map (star) (D) .

Uterine Endometrium
The prognosis of endometrial cancer depends on various factors, including histologic subtype and grade and the presence of lymph node metastases. Endometrial cancer is usually demonstrated on T2-weighted MRI; however, conventional MRI does not always demonstrate the tumor focus because the signal is variable and sometimes indistinguishable from normal endometrium or adjacent myometrium. 14 The addition of dynamic contrast-enhanced MRI results in a diagnostic accuracy of 85% to 93%. It can be difficult, however, to detect cancer on nonenhanced and contrast-enhanced MRIs in cases when an endometrium is of normal thickness or when adenomyosis obscures the tumor and invasion into the myometrium. The ADC value of endometrial cancer is significantly lower than that of endometrial polyps and normal endometrium. Furthermore, the ADC values in endometrial cancer of higher grades tend to be decreased compared with those of lower grades.

Uterine Myometrium
Malignant tumors of the myometrium consist of leiomyosarcoma, which is the most common, followed by endometrial stromal sarcoma. Leiomyosarcoma is frequently misdiagnosed as benign leiomyoma with an unusual degree of genital bleeding. Because an endometrial biopsy is usually not helpful for the definitive diagnosis of uterine sarcomas, MRI may play an important role in the diagnosis. Uterine myometrial sarcomas often show intermediate to high signal intensity on T2-weighted MRIs. Leiomyomas may occasionally be associated with various types of degeneration or cellular histologic subtype, which can cause increased signal on T2-weighted images. In such instances, the differentiation between benign and malignant myometrial tumors may be difficult if only based on signal intensity of nonenhanced and postcontrast MRI sequences. For such lesions, DW imaging is helpful in differentiating uterine sarcomas from leiomyomas by showing restricted diffusion in the former, whereas ordinary leiomyomas (including degenerating leiomyomas, which demonstrate high signal intensity on T2-weighted MRIs) do not reveal restricted diffusion. 15 Both uterine sarcomas and cellular leiomyomas exhibit high signal intensity on DW images, whereas ordinary leiomyomas and most degenerated leiomyomas show low signal intensity. Still, ADC measurement may have a limited role because of a large overlap between sarcomas and benign leiomyomas.
The role of DW imaging in the assessment of treatment response after uterine artery embolization is still under investigation. Preliminary results demonstrated a significant decrease of the ADC value after uterine leiomyoma treatment. Perfusion MRI imaging, however, is still the gold standard for follow-up treatment because of the direct information of the lesion vascularity.

Ovarian Tumors
DW imaging has been shown to demonstrate high signal intensity not only in the primary site of the ovarian cancer but also in disseminated peritoneal implants. 5 DW imaging has been shown to detect peritoneal implants with high sensitivity compared with T2-weighted MRIs ( Figure 6-4 ). Among malignant ovarian tumors, the ADC varies widely, probably as a result of variety in cellular morphology. In cystic lesions the ADC value may be mainly influenced by the viscosity of the fluid. Decreased ADC values have been reported in benign endometrial cysts, mature cystic teratomas, and some malignant cystic ovarian tumors. The cystic components of mature cystic teratomas had significantly lower ADC values than endometrial cysts, malignant neoplasms, and benign lesions. Detecting the keratinoid substance by DW imaging and the ADC value may be useful and serve as an adjunctive tool to ensure the accuracy of the diagnosis, particularly in patients with a mature cystic teratoma that lacks visible fat. No significant difference in the ADC value has been seen between benign and malignant cystic neoplasms. Because endometrial cysts tend to contain blood and some hemosiderin, the T1 values are shortened, resulting in a decrease in the ADC. Thus the potential usefulness in the differential diagnosis of cystic lesions remains controversial ( Figure 6-5 ).

FIGURE 6-4 Stage IVA cervical cancer. Axial T2-weighted turbo spin echo (A) and sagittal fat-suppressed T1-weighted postcontrast (B) magnetic resonance images (MRIs) in an older patient with cervical carcinoma demonstrate a mass with parametrial invasion on the ventral side (arrow) . A fat line between the tumor and the bladder wall cannot be appreciated. Enhancement of the dorsal bladder wall can be seen. The tumor is easily appreciated as a mass of high signal intensity on axial diffusion MRI ( b = 800) (D) with a corresponding area of restricted diffusion on the apparent diffusion coefficient map (star) (C) . The diffusion-weighted MRI shows high invasion of the bladder wall.

FIGURE 6-5 Ovarian cancer recurrence. Axial T2-weighted turbo spin echo magnetic resonance image (A) in a patient with increasing cancer antigen-125 level shows a lesion with high (cystic) (star) and low signal intensity (solid) (arrow) in the pouch of Douglas. The solid deposit is better appreciated on the apparent diffusion coefficient (B) image.

Lymph Node Metastasis
The presence of lymph node metastases is an important issue for patients with gynecologic cancers because it influences the 5-year survival rate and affects patient management. Both benign and malignant lymph nodes demonstrate high signal intensity at high b -values (i.e., >1000 s/mm 2 ). The highly cellular tissue in reactive lymph nodes may show increased signal intensity. The presence of necrosis within a lymph node represents a potential pitfall in lymph node analysis. The combination of size criteria (short axis diameter ≥5 mm with long axis diameter ≥11 mm, or short axis-to-long axis ratio >6 mm) and relative ADC values (<0.10 × 10 −3 mm 2 /s) resulted in a sensitivity and specificity of 83% and 99%, respectively. The latter method, however, used a combination of four complicated parameters (ADC of primary tumor, ADC of lymph node, and short and long axis diameters or ratio of lymph nodes) and may not be easily incorporated into an MRI protocol.

Bone Metastasis
The highly cellular tissue in red bone marrow may show increased signal intensity. The utility of DW imaging for the detection of bone metastases has not been explored.

Image Interpretation

A three-dimensional display of images acquired in two dimensions with a reversed gray scale can produce positron emission tomography-like images because of strong background suppression with excellent contrast. 5 This may cause difficulties, however, in correlating DW imaging findings with anatomy. Fusion software, which automatically overlays DW images onto anatomic MRIs, may resolve the poor anatomic information of the DW sequence. Obtaining DW and anatomic MRI sequences with uniform slice thickness, interslice gap, and field of view may also overcome the correlation problems.


DW imaging provides quantification of brownian motion of water protons by calculating the ADC, and can be used for in vivo quantification of the combined effects of capillary perfusion and diffusion.
In the female pelvis, DW imaging can be applied for tumor detection, to differentiate between benign and malignant lesions, and for the detection of cancerous lymph nodes.
The ADC of cervical cancers is significantly lower than the median ADC of normal cervix stroma.


1. Koyama T., Togashi K. Functional MR imaging of the female pelvis. J Magn Reson Imaging . 2007;25:1101-1112.
2. Le Bihan D. Diffusion/perfusion MR imaging of the brain: from structure to function. Radiology . 1990;177:328-329.
3. Namimoto T., Awai K., Nakaura T., et al. Role of diffusion-weighted imaging in the diagnosis of gynecologic diseases. Eur Radiol . 2009;19:745-760.
4. Nasu K., Kuroki Y., Nawano S., et al. Hepatic metastases: diffusion-weighted sensitivity-encoding versus SPIO-enhanced MR imaging. Radiology . 2006;239:122-130.
5. Takahara T., Imai Y., Yamashita T., et al. Diffusion weighted whole body imaging with background body signal suppression (DWIBS): technical improvement using free breathing, STIR and high resolution 3D display. Radiat Med . 2004;22:275-282.
6. Bammer R., Skare S., Newbould R., et al. Foundations of advanced magnetic resonance imaging. NeuroRx . 2005;2:167-195.
7. Stehling M.K., Turner R., Mansfield P. Echo-planar imaging-magnetic resonance imaging in a fraction of a second. Science . 1991;254:43-50.
8. Feinberg D.A., Jakab P.D. Tissue perfusion in humans studied by Fourier velocity distribution, line scan, and echo-planar imaging. Magn Reson Med . 2003;49:177-182.
9. Bastin M.E., Le Roux P. On the application of a non-CPMG single-shot fast spin-echo sequence to diffusion tensor MRI of the human brain. Magn Reson Med . 2002;48:6-14.
10. Pipe J.G., Farthing V.G., Forbes K.P. Multishot diffusion-weighted FSE using PROPELLOR MRI. Magn Reson Med . 2002;47:42-52.
11. Li T.Q., Takahashi A.M., Hindmarsh T., et al. ADC mapping by means of a single-shot spiral MRI technique with application in acute cerebral ischemia. Magn Reson Med . 1999;41:143-147.
12. Edelman R.R., Wielopolski P., Schmitt F. Echo-planar MR imaging. Radiology . 1994;41:143-147.
13. Basser P.J. Inferring microstructural features and the physiological state of tissues from diffusion-weighted images. NMR Biomed . 1995;8:333-344.
14. Kinkel K. Pitfalls in staging uterine neoplasm with imaging: a review. Abdom Imaging . 2006;31:164-173.
15. Tamai K., Koyoma T., Saga T., et al. The utility of diffusion-weighted MR imaging for differentiating uterine sarcomas from benign leiomyomas. Eur Radiol . 2008;18:723-730.

Suggested Readings

Namimoto T. Role of diffusion-weighted imaging in the diagnosis of gynecological diseases. Eur Radiol . 2009;19:745-760.
Qayyum A. Diffusion-weighted imaging in the abdomen and pelvis: concepts and applications. Radiographics . 2009;29:1797-1810.
Part Four
Chapter 7 Hysterosalpingography
Techniques, Normal Anatomy, and Pitfalls

Lauren M. Brubaker, Richard L. Clark
First introduced in 1910, hysterosalpingography (HSG) has more recently been eclipsed by the development of newer methods of imaging the female genital tract. Techniques such as endovaginal ultrasound and cross-sectional imaging, including computed tomography (CT) and magnetic resonance imaging (MRI), have now become standard for evaluating abnormalities of the female pelvis. Thus the use of HSG has shifted from the diagnosis of female pelvic disease to the evaluation of infertility. Despite the popularity of the newer imaging modalities, with the increasing rate of infertility (7.3 million women or 12% of the reproductive age population) 1 in the United States, the number of HSGs performed per year has gradually increased.


In evaluation for infertility, HSG is typically used to study the internal luminal morphology of the endocervical canal, uterine cavity, and fallopian tubes and their associated abnormalities, such as congenital anomalies, neoplasia, and inflammatory changes ( Figure 7-1 ). Within the uterus these include distortion of the endocervical canal secondary to myomas and identification of endoluminal masses such as polyps, myomas, and hyperplasia. Changes secondary to diethylstilbestrol (DES) exposure in utero can also be identified; however, this cohort of patients has now passed childbearing age. In addition to confirming tubal patency, HSG is useful in the diagnosis of various causes of tubal occlusion. These are most commonly the sequelae of pelvic inflammatory disease but also can be caused by obstructing polyps and endometriomas. Finally, HSG is used to confirm operative sterilization and to plan for reversal ( Figures 7-2 and 7-3 ). With the use of oil-soluble contrast media and selective tubal catheterization, HSG also has a potential therapeutic role in increasing the probability of pregnancy ( Box 7-1 ). 2 - 6

FIGURE 7-1 Duplication anomaly: bicornuate versus septate uterus. With hysterosalpingography, it is difficult to differentiate between a bicornuate and septate uterus as in this example. Rotation of the patient allows for free spill of the contrast media, which may outline the external contour of the uterus and allow for differentiation. These patients, however, frequently require three-dimensional ultrasound or magnetic resonance imaging to definitively differentiate between these entities.

FIGURE 7-2 Essure (Conceptus, San Carlos, CA) is an intra–fallopian tube device used for sterility. After its placement, hysterosalpingography is used to ensure appropriate occlusion of the fallopian tubes. In this example, there is no evidence of contrast media extending distal to the Essure devices, confirming successful occlusion of the fallopian tubes.

FIGURE 7-3 Tubal ligation. Note the truncated, slightly dilated fallopian tubes. To reverse tubal ligation, approximately 2 cm of fallopian tube are required for successful anastomosis. Mild tubal dilation is a normal finding after tubal ligation. 24

BOX 7-1 Indications

Indications for Hysterosalpingography

Confirmation of tubal patency in infertile patients
Confirmation or assessment for reversal of sterilization procedures
Filling defects within the endometrial canal

Although the contraindications for HSG are few, there are some instances where it cannot be performed safely ( Table 7-1 ). Active pelvic infection is considered an absolute contraindication because the retrograde injection of contrast material may cause further spread of the infection. Active vaginal bleeding is also an absolute contraindication because of the possibility of flushing clots into the peritoneal cavity, a condition that could lead to infection or endometriosis. 6, 7 Recent uterine or tubal surgery and uterine curettage are relative contraindications given the increased risk of intravasation. In these circumstances, HSG should not be performed for approximately 6 weeks to allow for adequate healing.
TABLE 7-1 Contraindications Absolute Relative Active pelvic infection Contrast media allergy Active vaginal bleeding Recent surgery Pregnancy  
As expected, normal intrauterine pregnancy is an absolute contraindication. Many examiners use the “10-day rule,” which states that an HSG should not be performed if it has been longer than 10 to 12 days between the onset of menses and day of procedure; if the patient has prolonged cycles (>28 days), this time may be extended up to 13 to 15 days. If there is any question of pregnancy, a pregnancy test should be obtained before the examination.
Finally, when there is a history of contrast agent allergy, the risks versus benefits of the procedure should be weighed. To our knowledge, there have been no known deaths from a contrast allergy during HSG; however, delayed reactions such as urticaria and hypotension have been reported. 8 In the case of a previous allergy to iodinated contrast media, the patient may be premedicated to help prevent allergic reaction. The American College of Radiology guidelines are 50 mg of prednisone orally 13 hours, 7 hours, and 1 hour before contrast administration, in addition to 50 mg of diphenhydramine orally 1 hour before the study. 9 The use of low osmolar contrast agents significantly decreases the risk of reactions to the administration of contrast agents. 10 Alternatively, one could consider using an alternative noniodine-containing contrast medium, such as gadolinium. 11


Injection Devices
Although many different HSG injection devices have been developed, there are three main types in use today. The choice of which device to use is primarily based on the individual examiner’s preference. The three most popular injection devices are the metal cannula, the cervical vacuum cup, and the balloon catheter ( Figure 7-4 ).

FIGURE 7-4 Injection devices: the three most popular injection devices for hysterosalpingography. A, Conventional metal cannula with a plastic acorn tip and separate metal tenaculum. B, Cervical vacuum cup. C, Balloon catheter with placement sheath.
The metal cannula is the historic preference for many radiologists and gynecologists. It is a rigid cannula with a plastic acorn tip, of varying sizes, to occlude the cervical os. A tenaculum is used to stabilize the cervix with countertraction while the cannula is inserted into the os. This technique has the advantage of giving one the ability to manipulate the uterus, which allows repositioning for optimal visualization of an anteverted or retroverted uterus. The use of the tenaculum, however, can cause pain, bleeding, and cervical laceration. For this reason, the cervical vacuum cup and balloon catheter techniques were developed, each of which offers various advantages and disadvantages over the metal cannula.
The cervical vacuum cup is a plastic device composed of a cervical adaptor with two injection tubes (one for injection and one for vacuum). It uses suction to adhere to the cervix and hold the central cannula in the cervical os during injection. The catheter comes in three cup sizes: 25-, 27-, and 30-mm diameters. This device has the advantage of decreased pain and ease of use; however, it does not allow easy manipulation of the uterus during the study.
The balloon catheter is currently the most popular injection device; it is composed of a soft, flexible cannula, available in 5- and 7-French sizes. The catheter is placed through the cervical os. Catheterization is facilitated by passing the catheter through a short, rigid plastic sheath. The balloon is then slowly inflated in the lower uterine segment to prevent inadvertent removal of the cannula during the examination. By using a flexible catheter instead of the rigid cannula, patient discomfort is reduced. One potential drawback, however, is that the inflated balloon can obscure the lower uterine segment. By inflating the balloon in the cervical canal, visualization of the lower uterine segment can be improved; however, this may increase pain or the frequency of vasovagal reaction. An alternative technique is to inflate the balloon in the uterus, inject contrast material until the fallopian tubes are visualized, then deflate the balloon, and inject additional contrast to better visualize the lower uterine segment and endocervical canal as the catheter is removed. Similar to the cervical vacuum cup, the balloon catheter does not allow the examiner to apply traction to the uterus for optimal visualization. Oblique images will help with this technical limitation.
Multiple recent studies have compared the metal cannula with the newer injection devices for ease of use, fluoroscopic time, and patient discomfort. The use of the cervical vacuum cup has been reported to decrease procedure and fluoroscopic time, pain, and volume of contrast agent compared with the metal cannula. 12 Similar studies comparing the use of metal cannulas with balloon catheters reported that the use of the latter also results in decreased fluoroscopic time, pain, and amount of contrast agent used. 13 When the cervical vacuum cup was compared with the balloon catheter, patients better tolerated the use of the balloon catheter over the cervical cup; however, the use of the cervical cup resulted in reduced radiation. 14

Contrast Media
There are two main classes of contrast media that can be used for HSG: water soluble and oil soluble. Oil-soluble agents have fallen out of favor in part because of the theoretical risk of clinically significant oil droplet emboli migrating to the lungs and, more importantly, the brain (right to left shunting). 15 Although rarely reported with earlier, more viscous oily agents such as Lipiodol (Guerbet, Villepinte, France), we are unaware of such complications with the currently available less viscous oily agent Ethiodol (Guerbet).
Virtually any water-soluble contrast agent containing approximately 300 to 350 mg I/mL can be used. These include meglumine diatrizoate, iopamidol, and iohexol. Occasionally a small amount of oil is administered via the uterine catheter after the HSG as a potential aid to fertility. 6

Technique Description
HSG is often performed as a joint effort between the Division of Reproductive Endocrinology, Department of Obstetrics and Gynecology, and the Department of Radiology. It is recommended that patients take a nonsteroidal antiinflammatory analgesic 1 hour before the procedure because pain and pelvic cramping are the main side effects of HSG. Before starting the procedure, the patient is asked to empty her bladder. The patient then lies supine on the fluoroscopic table in the lithotomy position. A sterile double-bladed metal speculum is inserted into the vagina. Alternatively a plastic speculum may be used and offers the advantage of being radiolucent. Once the cervix is exposed, povidone–iodine is used to prep the cervix. Any excess povidone–iodine should be wiped away because it can act as an irritant.
A scout radiograph is typically not necessary because clinically significant pelvic calcifications or other densities will be visualized with fluoroscopy before contrast administration. Contrast medium is then slowly injected while imaging under intermittent fluoroscopy. At least three to five spot radiographic films are recommended. 16 An anteroposterior spot radiograph is obtained during early uterine filling to detect small endometrial lesions that might be obscured by larger amounts of contrast media. Next a spot radiograph documenting complete filling of the uterus and early filling of the fallopian tubes before peritoneal spill is taken. Finally a spot radiograph demonstrating free bilateral peritoneal spill is taken to document tubal patency. Bilateral oblique spot radiographs may also be obtained for improved assessment of uterine pathology and for optimal visualization of the fallopian tubes. As described earlier, when using the balloon catheter, a postdeflation radiograph should be obtained to better evaluate the lower uterine segment and endocervical canal. With any injection device, if endometrial or endocervical pathology is suspected, postdrainage films should be obtained.
Once there is evidence of free intraperitoneal spill, contrast administration is discontinued. Typical volumes of administered contrast media range from 5 to 10 mL, but it is not unusual to instill 20 to 50 mL if the uterine cavity is enlarged secondary to myomas or previous pregnancies. It is important to use adequate volumes of contrast media to definitively visualize free intraperitoneal extension of contrast media ( Figure 7-5 ). Conversely, administration of an excessive amount of contrast media into a hydrosalpinx increases the risk of tuboovarian abscess; thus care is necessary. If at any time during the examination there is evidence of venous, lymphatic, or interstitial myometrial intravasation, contrast medium administration is immediately discontinued secondary to the increased risk of infection and pain ( Figure 7-6 ).

FIGURE 7-5 Hydrosalpinx. Pelvic inflammatory disease is a common cause of tubal disease. Most commonly there is occlusion of the fallopian tube secondary to adhesions, which causes dilatation of the fallopian tube or hydrosalpinx. The dilated fallopian tube is believed to be toxic to successful implantation, and thus hydrosalpinges are surgically removed before assisted reproductive treatment.

FIGURE 7-6 Intravasation. After administration of contrast media, intravasation of contrast was identified within the myometrium of the uterus, characteristic of venous and probable lymphatic intravasation.

Normal Anatomy
The normal uterine cavity is pyramidal in shape with the apex located inferiorly and pointing to the endocervical canal ( Figure 7-7 ). The uterine fundus, or base of the pyramid, is usually straight; however, it can have a slightly concave or arcuate configuration. The uterus is typically in an anteverted position, with retroversion less common ( Figure 7-8 ). During the proliferative phase, the contour of the internal endometrial cavity is usually smooth. Uterine ridging, or longitudinal folds within the uterine cavity that parallel the long axis of the uterus, may be visualized and represent a normal variant of the endometrial pattern. Within the first 4 days after menses, the uterine cavity may have an irregular or shaggy appearance, which is a normal finding and must not be mistaken for endometrial pathology ( Figure 7-9 ). Occasionally there is extension of contrast media into the wall of the uterus, indicating adenomyosis ( Figure 7-10 ). Large, smooth, and symmetric indentations may occasionally be visualized from uterine contractions during the examination. Intrauterine filling defects commonly represent endometrial polyps or leiomyomas ( Figure 7-11 ).

FIGURE 7-7 Normal anatomy ( U , uterus; short arrow , cornua of the fallopian tube; long arrow , isthmus of the fallopian tube; curved arrow , ampulla; arrowhead , cervical canal with catheter in place).

FIGURE 7-8 Anteverted uterus with the patient in a right posterior oblique position (A) and left posterior oblique position (B) B, Retroverted uterus with the patient in a right posterior oblique position (C) and left posterior oblique position (D) . To determine whether the uterus is anteverted or retroverted, oblique radiographs are helpful.

FIGURE 7-9 Irregular uterine wall. Within the first 4 days of the patient’s menses, the uterine wall may have an irregular or shaggy appearance such as seen above. This is a normal finding and should not be confused with endometrial pathology.

FIGURE 7-10 Adenomyosis. Adenomyosis, or small endometrial diverticula extending into the myometrium, can be focal or diffuse as in this example. This condition usually manifests as abnormal uterine bleeding or pelvic pain and thus may be an incidental finding on HSG.

FIGURE 7-11 Intrauterine filling defects. Two large filling defects are noted within the uterine cavity, which persist with contrast filling of the uterus. These filling defects represent two large endometrial polyps. Submucosal or pedunculated leiomyomas may have a similar appearance and should be differentiated.
The normal endocervical canal varies in length and diameter. The endocervical mucosa may have multiple, small 5- to 10-mm parallel folds, which correspond to normal endocervical glands (plicae palmatae).
The fallopian tubes are thin, long tubular structures and typically measure approximately 10 to 12 cm in length. They can be radiographically divided into three segments: the short interstitial or cornual segment, isthmus, and ampulla. The interstitial portion frequently appears triangular in shape where it enters the uterus. The isthmus is the long, narrow portion with an approximately 2-mm lumen and a thick muscular wall. It begins to dilate just before its junction with the ampullary region. The ampulla has a thinner muscular wall and increased diameter (approximately 5 mm) and terminates at the fimbria, where contrast media may occasionally be seen to outline the ovary. Rugal folds are nearly always visualized in the ampullary region, and in fact their absence usually indicates chronic inflammation. Typically the fallopian tubes curve posterolaterally and can often be better visualized in their entirety on oblique spot radiographs.

Technical Problems and Solutions

Inadequate Visualization
Visualization of the uterus and fallopian tubes can be improved with simple variations in technique.
• If the uterus is difficult to visualize when using a balloon catheter, have the patient perform a Valsalva maneuver. 17 This will alter the position of the uterus for better visualization. Oblique radiographs may also be useful for improved visualization.
• When there is no evidence of contrast flow into the fallopian tubes and cornual spasm is suspected, patient relaxation and the “tincture of time” usually result in reversal of the spasm ( Figure 7-12 ).
• When a septate or bicornuate uterus is suspected, after free spill of the contrast media, rotate the patient 360 degrees or place her prone to outline the external contour of the uterus with contrast media. Knowing the external contour of the uterus may help differentiate between these two entities. Frequently, however, these patients will go on to three-dimensional ultrasound or MRI for further characterization (see Figure 7-1 ).
• For optimal visualization of pelvic pathology and tortuous fallopian tubes, oblique radiographs are invaluable.
• To differentiate between air bubbles and polyps, a small volume of contrast media can be flushed into the uterine cavity and then aspirated back into the syringe. After this maneuver, a polyp will remain fixed, whereas an air bubble will move. Similarly, review of early injection radiographs or delayed postdrainage radiographs may also help differentiate air bubbles from polyps ( Figure 7-13 ).

FIGURE 7-12 Cornual spasm. A, Initially there is no evidence of contrast filling the right fallopian tube. B, With time and patient relaxation, there may be reversal of the cornual spasm and normal contrast filling of the fallopian tube as in this example.

FIGURE 7-13 Air bubbles. A, Two filling defects are identified in the right cornua. B, With further administration of contrast media, these filling defects are no longer visualized, compatible with air bubbles. Persistent small filling defect in the left cornua is consistent with a small polyp. Incidental note of a right hydrosalpinx is made.

Complications/Side Effects
The most frequent side effect of HSG is pain ( Box 7-2 ). Although typically short-lived, up to 80% of patients report mild to moderate pain during the procedure. 18 In the past, HSG-induced pain was thought to be secondary to irritation of the peritoneal cavity from high-osmolality contrast media; however, this theory was disproved when there was no difference in the prevalence of pain after use of high- or low-osmolar contrast agents. 19 The prevalence of pain also differs with the injection device and examiner’s technique. For example, pain is decreased with the use of a balloon catheter or cervical vacuum cup compared with the metal cannula. Pain is also usually decreased with slow inflation of the balloon or application of the vacuum, the use of topical anesthesia (benzocaine gel), and slow injection of contrast medium. 8

BOX 7-2 Complications

Complications of Hysterosalpingography

Lymphatic or venous intravasation
Pelvic infection
Radiation exposure
Allergic reaction to contrast media
Vasovagal reaction
Uterine perforation
Another complication is lymphatic and venous intravasation, which occurs in up to 7% of patients. 20 Although intravasation can be seen in patients with normal HSG, patients with tubal obstruction are at higher risk. Other risk factors for intravasation include recent uterine surgery, uterine scarring, aberrant placement of the cannula, excessive pressure of injection, and the presence of fibroids. Radiographically, venous intravasation is characterized by contrast-filling multiple, parallel, or tortuous channels that course cephalad, such as within the ovarian vein. In myometrial intravasation, contrast medium fills the interstitial microvascular network of the uterine wall (see Figure 7-6 ). With lymphatic intravasation, contrast may occasionally be seen filling larger lymphatics, which are characterized by their typical beaded appearance.
Venous or lymphatic intravasation poses several risks, the most important being pelvic infection, either new or exacerbated. Even in the absence of contrast intravasation, pelvic infection can be seen in 0.3% to 3.1% of hysterosalpingograms. The most important contributing risk factor is a hydrosalpinx or other tubal obstruction. 20 In most cases, the etiology of the pelvic infection after HSG is likely the result of the introduction of cervical organisms during the procedure, emphasizing the need for appropriate sterile technique. In cases of pelvic infection, patients usually present with fever and pelvic pain within 12 to 48 hours after HSG. Before the procedure all patients should be informed about the symptoms of pelvic infection because early treatment is imperative; untreated infection can lead to tuboovarian abscess, sepsis, or even death. In the presence of a newly diagnosed hydrosalpinx or other tubal obstruction, patients should be treated prophylactically with 200 mg of doxycycline orally before leaving the department, followed by 100 mg orally twice daily for 5 days. Patients with preexisting tubal obstruction should receive antibiotics before the examination.
Examiners should always be aware of the radiation exposure associated with HSG. According to the literature, for a typical HSG examination, the equivalent dose to the ovaries is approximately 4.5 mSv, and the equivalent dose to the uterus is approximately 3.5 mSv. 21 Because the majority of these studies are performed in women of childbearing age, it is imperative to use as little radiation as needed to perform the study. Dose-limiting techniques include limiting the field of view and appropriate shielding. In addition, minimizing fluoroscopic time and limiting the number of spot radiographs will also reduce the overall radiation dose. The majority of HSGs can be completed with less than 2 minutes of fluoroscopy time.
Allergic reactions to iodinated contrast agents, although extremely rare, can occur. If clinically necessary, steroid premedication can be used per American College of Radiology guidelines as outlined above. Vasovagal reactions with hypotension and bradycardia can occur with inflation of the balloon catheter in the uterine cavity or endocervical canal. These usually resolve quickly with removal of the catheter or balloon deflation. Finally, there is the risk of perforation of the uterus or fallopian tubes with the cannula or catheter. The risk of perforation is decreased by using flexible plastic catheters and careful technique. When a perforation does occur, it is usually treated conservatively.

Diagnostic Accuracy
Multiple studies have compared the accuracy of HSG with that of laparoscopy and laparotomy in the diagnosis of tubal disease. For tubal occlusion the overall false-positive rate of HSG is estimated to be 6% to 29%, largely secondary to insufficient administration of contrast media or unrecognized spasm. The false-negative rate is between 4% and 7%, accounted for by undetected peritubal adhesions. 22 False-negative results can also be secondary to incomplete filling of a dilated hydrosalpinx ( Figure 7-14 ). Identification of peritubal adhesions in the presence of a hydrosalpinx can be particularly difficult, with a reported 44% false-negative rate compared with laparoscopy ( Figure 7-15 ). 23 Thus in patients in whom HSG is equivocal or negative and the clinical suspicion remains high, laparoscopy may be indicated.

FIGURE 7-14 The danger of incomplete filling of the fallopian tubes. A, With early filling, the left fallopian tube appears normal with seemingly early spill of free intraperitoneal contrast. B, With additional administration of contrast media, there is obvious dilatation of the left fallopian tube, consistent with a left hydrosalpinx. This finding would have been missed with incomplete filling.

FIGURE 7-15 Peritubal adhesive disease. Note the loculations of contrast media in the peritoneum as opposed to free spill. This appearance is compatible with bilateral peritubal adhesive disease, which is frequently secondary to pelvic inflammatory disease.


HSG allows for visualization of the internal morphology of the endocervical canal, uterine cavity, and fallopian tubes.
HSG is most commonly used in the evaluation of infertility.
Currently there are three devices to safely perform HSG; the most favored is the balloon catheter.
Although the most frequent complication of HSG is pain, the most worrisome is pelvic infection.
Doxycycline is commonly administered to patients to prevent infectious complications after HSG when there is venous or lymphatic intravasation or tubal occlusion.
Optimizing fluoroscopic technique allows for better visualization and will usually result in a diagnostic study.
Allergic reactions to iodinated contrast agents, although extremely rare, can occur.


1. Assisted Reproductive Technology Success Rates 2006. US Department of Health and Human Services and the Center for Disease Control and Prevention. . Accessed date 5/20/09.
2. Watson A., Vandekerckhove P., Lilford R., et al. A meta-analysis of the therapeutic role of oil soluble contrast media at hysterosalpingography: a surprising result? Fertil Steril . 1994;61:470-477.
3. Rasmussen F., Lindequist S., Larsen C., et al. Therapeutic effect of hysterosalpingography: oil- versus water-soluble contrast media—a randomized prospective study. Radiology . 1992;185:283-284.
4. Johnson N., Vandekerckhove P., Watson A., et al. Tubal flushing for subfertility. Cochrane Database Syst Rev . 2005;2:CD003718.
5. Luttjeboer F., Harada T., Hughes E., et al. Tubal flushing for subfertility. Cochrane Database Syst Rev . 2007;3:CD003718.
6. Steiner A.Z., Meyer W.R., Clark R.L., et al. Oil-soluble contrast during hysterosalpingography in women with proven tubal patency. Obstet Gynecol . 2003;101:109-113.
7. Berek J., Novak E. Berek & Novak’s gynecology , ed 14. Philadelphia: Lippincott Williams & Wilkins; 2006.
8. Katzberg R.W. The contrast media manual . Baltimore: Williams & Wilkins; 1992.
9. ACR Contrast Media Guidelines. . Accessed 5/20/09.
10. Katayama H., Yamaguchi K., Kozuka T. Adverse reactions to ionic and nonionic contrast media. Radiology . 1990;175:621-628.
11. De Caro G., Ferraiolo A., Castelletti L., et al. Hysterosalpingography with gadolinium. Eur Radiol . 2005;15:1469-1471.
12. Cohen S.B., Wattiez A., Seidman D.S., et al. Comparison of cervical vacuum cup cannula with metal cannula for hysterosalpingography. BJOG . 2001;108:1031-1035.
13. Tur-Kaspa I., Seidman D.S., Soriano D., et al. Hysterosalpingography with a balloon catheter versus a metal cannula: a prospective, randomized, blinded comparative study. Hum Reprod . 1998;13:75-77.
14. Ricci G., Guastalla P., Ammar L., et al. Balloon catheter vs. cervical vacuum cup for hysterosalpingography: a prospective, randomized, single-blinded study. Fertil Steril . 2007;87:1458-1467.
15. Lindequist S., Justesen P., Larsen C., et al. Diagnostic quality and complications of hysterosalpingography: oil- versus water-soluble contrast media—a randomized prospective study. Radiology . 1991;179:69-74.
16. ACR Guidelines for Techniques: Hysterosalpingograhy. . Accessed 5/20/09.
17. Lindheim S.R., Sprague C., Winter T.C.3rd. Hysterosalpingography and sonohysterography: lessons in technique. AJR Am J Roentgenol . 2006;186:24-29.
18. San Fillippo J.S., Yussman M.A., Smith D. Hysterosalpingography in the evaluation of infertility: a six year review. Fertil Steril . 1978;30:636.
19. Winfield A.C., Wentz A.C. Diagnostic imaging of infertility , ed 2. Baltimore: Williams & Wilkins; 1992.
20. Yoder I.C. Hysterosalpingography and pelvic ultrasound: imaging in infertility and gynecology . Little Brown: Boston; 1988.
21. Kramer R., Khoury H.J., Lopes C., et al. Equivalent dose to organs and tissues in hysterosalpingography calculated with the FAX (Female Adult voXel) phantom. Br J Radiol . 2006;79:893-898.
22. Holst N., Abyholm T., Borgersen A. Hysterosalpingography in the evaluation of infertility. Acta Radiol Diagn (Stockh) . 1983;24:253-257.
23. Duff DE, Fried AM, Wilson EA, et al: Hysterosalpingography and laparoscopy: a comparative study, AJR Am J Roentgenol 198;141:761–763.
24. Kerner T.C., Clark R.L., Yankaskas B.C., et al. Post-ligation dilatation of the fallopian tube. Invest Radiol . 1984;19:213.

Suggested Readings

Ott D.J., Fayez J.A. Hysterosalpingography: A Text and Atlas . Baltimore-Munich: Urban & Schwarzenberg; 1991.
Winfield A.C., Wentz A.C. Diagnostic imaging of infertility , ed 2. Baltimore: Williams & Wilkins; 1992. pp 1-8
Yoder I.C. Hysterosalpingography and pelvic ultrasound: imaging in infertility and gynecology . Boston/Toronto: Little, Brown and Company; 1988.
Section Two
Pelvic Pain
Chapter 8 Approach to Pelvic Pain and the Role of Imaging

Rochelle F. Andreotti, Geoffrey E. Wile, Sara M. Harvey
Pelvic pain is a frequent problem for women and may account for 10% to 40% of all gynecologic visits. 1 It is the most common reason for laparoscopy in the United Kingdom, and in the United States, it is the second most common reason for laparoscopy and the most common reason for hysterectomies, accounting for 12% to 19% of all hysterectomies performed. 2, 3
Diagnostic considerations encompass gynecologic and obstetric causes as well as nongynecologic etiologies. Because the selection of imaging modality is determined by the clinically suspected differential diagnosis, a careful evaluation of the patient should be performed and diagnostic considerations narrowed before a modality is chosen.
Transvaginal ultrasound (TVUS) and transabdominal ultrasound (TAUS) of the pelvis have the ability to narrow the differential diagnosis and are the imaging modalities of choice when a gynecologic etiology is suspected. Sonography should also be considered when gastrointestinal or urinary tract pathology is suspected in the pregnant patient. Not only is it noninvasive, radiation free, and cost effective, but sonography also accurately delineates the architecture of the uterus and ovaries. Because of better anatomic resolution, TVUS should be used whenever possible, although a routine study usually includes both techniques, and TAUS provides more information when uterine and adnexal structures are beyond the field of view of the transvaginal probe. In addition, spectral and color or power Doppler imaging can be used to characterize vascularity to the ovaries, adnexal structures, and uterus, which may also be beneficial in narrowing the field of differential considerations. The majority of patients presenting with pelvic pain who have a normal pelvic ultrasound (US) are likely to have resolution of their pain, and further imaging is unlikely to yield positive results. 4 There are a few situations in which TVUS may be contraindicated. These include premenarchal patients, the majority of virginal patients, and any patient who does not willingly consent to a vaginal examination. The examination may need to be discontinued in the patient with a narrow introitus who experiences discomfort at the time of transducer insertion. In general, TVUS can be performed in any patient in whom a bimanual examination is appropriate.
Computed tomography (CT) is used more frequently when a gastrointestinal or genitourinary abnormality is likely, although gynecologic disease is being detected with increasing frequency by CT as a result of 24-hour accessibility in many hospitals and advances in CT technology leading to faster acquisition times and multiplanar reformats.
Magnetic resonance imaging (MRI) is usually favored over CT for assessing the pregnant patient because of the lack of ionizing radiation but may be hampered by its lack of widespread availability, especially in the acute setting. MRI is also emerging as a problem-solving tool for specific indications, if further characterization of a disorder is required and when US is found to be inadequate. For example, if the patient’s pain fails to resolve or is more chronic in nature, MRI is often beneficial in the evaluation of adenomyosis and endometriosis, which usually demonstrate characteristic MRI findings.
Discussions of pelvic pain generally distinguish acute pelvic pain from chronic pelvic pain. Acute pain is intense pain characterized by sudden onset. 5 Chronic pain is typically defined as noncyclic pelvic pain that lasts 6 months or longer and that is severe enough to cause functional disability or the need for medical care. 5 Although chronic pelvic pain may occur in as many as 15% of the population, 60% go undiagnosed, in contrast to patients with acute pain, in whom a diagnosis is more likely to be reached. 4 The time frame for the definition of chronic pain is arbitrary, and alternative durations such as 3 months have been suggested. 6 In addition to acute and chronic pain, some pain is cyclic, referring to pain that is associated with the menstrual cycle. Dysmenorrhea (i.e., painful menstruation) is the most common form of cyclic pain. 5
This chapter provides an overview and an approach to the evaluation of the female patient with acute pelvic pain, chronic pelvic pain, and dysmenorrhea. Imaging findings of the causes of pain discussed briefly here are discussed in more detail in other chapters, including specific chapters on acute and chronic pelvic pain ( Chapters 10 and 11 ).

  • Accueil Accueil
  • Univers Univers
  • Ebooks Ebooks
  • Livres audio Livres audio
  • Presse Presse
  • BD BD
  • Documents Documents