Gynecologic Imaging E-Book


1058 Pages
Read an excerpt
Gain access to the library to view online
Learn more


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


Published by
Published 05 April 2011
Reads 0
EAN13 9781437735987
Language English
Document size 6 MB

Legal information: rental price per page 0.0770€. This information is given for information only in accordance with current legislation.

Report a problem

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,
S a u n d e r sFront 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
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, OregonCopyright
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 : eld 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 identi: ed, 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
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
[DNLM: 1. Genitalia, Female–radiography. 2. Diagnostic Imaging–methods. 3.
Genital Diseases, Female–diagnosis. 4. Pregnancy Complications–radiography. WP
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 1Contributors
Lejla Aganovic, MD , Assistant Professor, Department of
Radiology, University of California, San Diego,
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,
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,
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,
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,
WashingtonChronic 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 ConceptionSteven 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,
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
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
Ovarian and Fallopian Tube CancerNancy 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
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,
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,
PennsylvaniaPitfalls 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
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
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,
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 PatientAndrew 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
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,
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, ConnecticutEndometriosis
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,
DartmouthHitchcock 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,
YaleNew 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,
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,
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 FistulasImaging 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 Pregnancy1
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 oor dysfunction. We have also included images of disease states beyond
the gynecologic tract that may mimic expected clinical ndings. 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 ndings.
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 di usion 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 broids, abscess drainage, and fallopian
tube canalization are presented, with expert advice on patient preparation,2
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 con dently 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 speci cally 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 speci c 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
Julia R. Fielding
Douglas L. Brown
Amy S. ThurmondDedication
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
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
Amy S. ThurmondTable 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
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
Section Two: Pelvic Pain
Chapter 8: Approach to Pelvic Pain and the Role of Imaging
Chapter 9: EndometriosisChapter 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
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 CancersChapter 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
Chapter 39: Uterine Artery Embolization
Chapter 40: Fallopian Tube Catheterization
Chapter 41: Ultrasound-Guided Treatment of Ectopic Pregnancy
Selections from the ACR Appropriateness Criteria
IndexSection One
Imaging Techniques, Pitfalls,
and Normal AnatomyPart One

Chapter 1
The Normal Pelvis on Ultrasound Imaging and
Anatomic Correlations
Tulin Ozcan, Vikram Dogra
Gynecologic imaging by ultrasound (US) has improved signi cantly 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 su, cient 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 e4ects in tissues it traverses: heating and
mechanical bioe4ects. The American Institute of Ultrasound in Medicine (AIUM)
guidelines conclude that there is no independently con rmed 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 rst trimester examinations and
1use of Doppler techniques.
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 7uid in the
cul-de-sac is a variant of the normal (Figure 1-1); however, increased amount of 7uid in
the posterior cul-de-sac and any 7uid in the anterior cul-de-sac usually indicates
FIGURE 1-1 Mild amount of free 7uid 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 7uid in the cavity can be seen secondary to high estrogen levels before
delivery. The prepubertal uterus has a tubular con guration; 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
2cm, and the anteroposterior dimension of the uterus does not usually exceed 10 mm.
The uterus starts to grow before menarche and continues to grow for several years.
The pubertal uterus has the adult pear con guration 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 xed 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
retro7exed 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 ante7exed uterus on transabdominal imaging in the sagittal

plane. B, Retroverted retro7exed 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 calci cations (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
con guration. The outer layer is thin and separated from the middle layer by arcuate
vessels, which are variably evident by US. Calci cations 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 7uid in the endometrial cavity, the endometrium
should be measured excluding the 7uid 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 broids, 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
3cornua) should be imaged in longitudinal and transverse views.

FIGURE 1-3 Transvaginal ultrasound in the sagittal plane demonstrates 7uid (arrow) in
the endometrial cavity of an 81-year-old patient. The endometrial measurements (calipers)
appropriately exclude the endometrial 7uid. 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-de ned endometrium gradually thickens up to approximately 8
mm. The functional and basal layers can be visually di4erentiated 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 in7uence 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, nally 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 signi cance 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
3when endometrial thickness is 4 mm or less. In the postmenopausal patient without any
bleeding, there is considerable disagreement for de ning the upper limit of endometrial

4,5thickness, giving a range of 5 to 15 mm. The signi cance 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
3,6for endometrial carcinoma using both hysteroscopy as well as dilation and curettage.
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
7specificity and with positive predictive values as low as 1.4%.
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 nding of nabothian cysts
(Figure 1-5). Nabothian cysts are usually of no clinical signi cance. 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 calci cations from
previous instrumentation. They generally are of no clinical significance.
FIGURE 1-5 Nabothian cyst (arrow) in the cervix on transvaginal sagittal ultrasound
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 nger-like projections, termed fimbriae, surround the
abdominal ostia of the tube. One of the largest mbriae 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 Wol, an 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 Wol, an duct or paramesonephric duct
remnants, with the latter occurring more commonly within the broad ligament rather
than at the mbriated 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
3 81 cm , and small cysts or follicles may be observed. The mean volume is approximately
3 31.1 cm among girls up to 1 year of age and decreases to 0.67 cm 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
8to 12 months. 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
3premenopausal women is approximately 7.4 to 7.8 cm (standard deviation [SD], 2.4 to
52.6) and is not in7uenced by parity. The ovarian volume decreases in the
postmenopausal patient showing a signi cant relationship with the number of years
postmenopause. Average ovarian volume in the early postmenopausal patient is 3.4 to
3 3 53.8 cm (SD, 1.3 to 1.6) and in late menopause is 2.5 cm (SD, 1.1 to 1.3).
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 di4er 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 speci city for follicle
9rupture. 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
rst 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
folliclestimulating 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 brin
strands with no blood 7ow 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 7uid 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, bromuscular 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 7attened
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
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 super cial transverse perineal muscle.
Posteriorly the urogenital diaphragm inserts into the central point of the perineum.
Situated farther posteriorly is the ischiorectal fossa. Located more super cially 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 bromuscular 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 bers of the endopelvic
fascia. The cardinal ligaments are composed of connective tissue that medially is united
rmly 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
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 o4 only two branches, the inferior
epigastric artery and the deep circum7ex 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 o4 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 7ow varies cyclically, with perfusion being
highest (and resistance to 7ow lowest) immediately before ovulation and in the midluteal
phase. In general, blood 7ow increases in response to estrogens. There appears to be no
di4erence in blood 7ow 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 di4erentiate
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 ll
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
7exed 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
MayerRokitansky-Küster-Hauser syndrome and vaginal septa, as well as assessment of structural
10and functional integrity of the pelvic floor.
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
• 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 broids 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 7uid 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 majoradvantage 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
11indications for a 3D examination in gynecology :
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 occlusionFIGURE 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 withheavy 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
12invasive diagnostic procedures. 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 7ow 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
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 7ushed 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 di, culty 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
13,14SHG in terms of agreement with hysteroscopy ndings. Procedure-related adverse
e4ects and complications are mild and uncommon. In a prospective study, the
complication rates were reported as failure to complete the procedure, 7%; pelvic pain,
153.8%; vagal symptoms, 3.5%; nausea, 1%; and postprocedure fever, 0.8%.
Preprocedure administration of nonsteroidal antiin7ammatory 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
Transabdominal and transvaginal routine images and measurements generally
include the following:
• Sagittal uterine views, with longitudinal and anteroposterior uterine
• Transverse Uterine views, with transverse uterine measurements at the funds.
• Cervical image in sagittal view revealing cul-de-sac and if necessary in transverse
• 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.
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.
7. Holbert T.R. Transvaginal ultrasonographic measurement of endometrial thickness in
postmenopausal women receiving estrogen replacement therapy. Am J Obstet Gynecol.
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.
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.
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 inMedicine 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 ndings 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 lling 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 broids, 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 su cient 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 eld 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 4exed 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 calci cations. Although most broids 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 signi cant
1cystic degeneration may simulate an ovarian cyst. 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 calci ed myoma
is usually not a diagnostic dilemma, but the shadowing may cause signi cant technical
limitations in evaluation of the adnexa and the endometrium.
Usually multiple broids 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 broid. Frequently the large broid is
mistaken for the entire uterus when this dominant broid displaces a smaller, more
normal-appearing uterine corpus toward the periphery (Figure 2-2). Distinguishing
between a solitary enlarged myoma and multiple myomata may a ect 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 broid mistaken as a di usely myomatous main body of
the uterus. A, Sagittal transabdominal ultrasound shows exophytic dominant broid
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.
Di use 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 de nition 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 broids 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 de ned margins, minimal mass e ect on the
adjacent endometrium, small cysts, ellipsoid versus spherical shape, presence of
echogenic foci and striations, and the absence of calci cations, edge shadowing, and!
2large vessels at the margin of the lesion. With the improvement of US equipment in the
past 10 to 15 years, we have seen several cases initially interpreted as broids that on
subsequent studies are focal regions of adenomyosis. Because patients with broids and
adenomyosis have similar symptoms, di erentiation between the two diagnoses is
important because treatment options di er. Finally, in cases with adenomyosis and
broids, shadowing from the broids may limit evaluation of the presence and severity of
the adenomyosis.
FIGURE 2-4 Focal adenomyosis in posterior uterine body myometrium confused for
broid. A, On transabdominal ultrasound image, a focal hypoechoic region (arrows) was
mistaken for broid. B, Transvaginal sagittal view of uterus con rms the area as focal
adenomyosis in light of its poor de nition (thick arrows), myometrial cyst (thin arrow), and
pencil-thin edge artifact. Cursors mark endometrium. Small anterior uterine body broid
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 di erentiate a
septate uterus from a bicornuate or didelphys uterus. Pregnancy outcomes and treatment
options vary signi cantly for these di erent anomalies. Volumetric US imaging is helpful
in di erentiating 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 broid 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 ndings of a broid should help distinguish these
two diagnoses.
False-Positive Diagnosis of Fibroids
Fibroids may calcify, but not all uterine myometrial calci cations are broids.
Calci cation of arcuate uterine arteries is common in elderly women, especially those
with diabetes, and is considered part of the normal aging process. Arcuate artery
calci cations usually appear as discontinuous linear parallel echoes (Figure 2-5) located!
3between the outer and intermediate layers of the uterine myometrium. These
calci cations must not be confused with calci cation in uterine broids, which tend to be
coarser and clumped and often associated with a noncalci ed mass. Intense shadows
from arcuate vessel calci cation may obscure the endometrium and make it di cult to
distinguish from the leading edge of a calci ed myoma. Magnetic resonance may be the
only imaging method for evaluation of the endometrium in this situation.
FIGURE 2-5 Calci cation of the uterine arcuate arteries. Sagittal transvaginal
ultrasound of the uterus reveals calci cations in a discontinuous parallel line
configuration (arrows) located in the mid to outer aspect of the uterine myometrium.
A frequent site for an erroneous broid 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
brous 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 broid
4(Figure 2-7). Thurmond et al. suggest that di erences 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
5endometrium above the scar, which may also account for this prominent tissue.!
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 di erent patients with cesarean section scars mistaken for broids. A,
Sagittal transvaginal ultrasound image reveals rounded con guration of the tissue
superior to the cesarean section scar (S). The arrow points to the bulbous tissue (mistaken
for a broid) located superior to the scar. B, Sagittal transvaginal ultrasound image
reveals edge artifact (thick arrow) originating from scar and rounded con guration 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 broid. This artifact manifests as narrow, vertical hypoechoic lines originating
6along the lateral margin of a rounded structure. Edge artifact is common along the
border of a cesarean section scar because of the rounded con guration of the
myometrium adjacent to the scar. The hypoechoic edge artifact may be mistaken for
posterior acoustic shadowing, further confusing this con guration for a broid (Figure
27, 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, broids, hyperplasia,
and cancer. Accurate identi cation of the endometrium may be challenging in patients
with multiple broids. Anterior broids that signi cantly attenuate the sound beam may
completely obscure the endometrium. In addition, one must avoid mistaking tissue
heterogeneity within broids or tissue interfaces between broids as the endometrium
(Figure 2-8).
FIGURE 2-8 Tissue interface between myometrial broids mistaken as the
endometrium. A, Cursors mark the erroneous measurement of the echogenic tissue
interface between two broids 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
7of reconstructed planes regardless of the original acquisition plane.
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 de nition 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-de ned endometrium in areas without
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
4postmenstrual spotting with old blood. This delayed bleeding results from either
5retained menstrual blood or in situ bleeding in an endometrial cavity pouch. 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!
FIGURE 2-11 Fluid within cesarean section scar niche. Transvaginal sagittal ultrasound
image of the cervix reveals a small amount of 4uid 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
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, broids, and in4ammation, 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.
Highfrequency 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 nd portions of the normal fallopian tubes, especially if
there is some free pelvic 4uid 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 mbria 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 nger-like projections of the mbria 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 di erent 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 mbriated end of the tube. This
fallopian tube is easily visualized because of adjacent free 4uid. 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 mbria (see Figure 2-12, A). Clinically they are insigni cant and rarely
symptomatic unless they undergo torsion and infarction.
When a fallopian tube dilates or becomes in4amed, it should be more easily identi ed.
Fallopian tube dilatation usually implies obstruction, although the converse is not
necessarily true. Although pelvic in4ammatory disease (PID) is the most common cause
of dilatation, other etiologies include endometriosis and adhesions from an in4ammatory
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
8endosalpingeal folds may be visualized. When the lumen occludes, the tube will dilate,
lling with complex 4uid 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
9demonstrate the “cogwheel sign.”
FIGURE 2-13 Pyosalpinx. Transvaginal ultrasound image reveals a thick-walled dilated
fallopian tube containing a 4uid/debris level of purulent material related to pelvic
in4ammatory disease. Note the “cogwheel sign” of the thickened endosalpingeal folds
In the chronic phase of PID, a hydrosalpinx may develop. The 4uid 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
10fallopian tube. 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 4uid 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 4uid- lled 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
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 con rmed with color Doppler US, although occasionally
due to slow 4ow, even spectral Doppler US will be negative. In these cases the
slowmoving echoes will be visible on gray scale, helping to di erentiate 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
FIGURE 2-15 Pelvic varices. A, Transvaginal ultrasound image reveals a dilated tubular
structure with ne internal echoes, and a 4uid/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 identi ed
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 con rm small bowel. Simple dilatation of small
and/or large bowel can be due to a primary process (obstruction, in4ammation,
ischemia) or a reactive ileus. US is useful to di erentiate normal and abnormal peristalsis
and may be the first clue to a bowel obstruction.
More focal abnormalities of bowel can suggest a speci c diagnosis. Thus a blind-ending
bowel loop larger than 6 mm may be identi ed 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-walledtubular structure with gut signature (Bowel) representing the in4amed 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
3number of follicles. The ovaries are often larger at birth with an average volume of 1 cm
and may be palpable as a result of multiple cysts, related to the in4uence of maternal
hormones (Figure 2-17). Such an appearance is not worrisome and will resolve during
3infancy. In infancy the ovarian volume is usually less than 1 cm , increasing gradually
3with age to 2 to 3 cm approaching menarche, continuing to increase in size during the
3 11teens with an average volume of 8 cm . 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
3women is reported as 9.8 ± 5.8 cm . After menopause the ovaries atrophy and follicles
disappear. Size is related to time since menopause with volumes varying between 1 and 5
3 3cm . Ovarian volume more than 10 cm in a postmenopausal woman is considered
12abnormal by some authors.
FIGURE 2-17 Normal newborn ovary. Transabdominal ultrasound image reveals
multiple small cysts in an enlarged ovary secondary to stimulation from maternal
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=
13common. 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
14small percentage as a result of chromosomal abnormalities. 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
signi cant di erential 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 signi cantly. 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 4ow in the a ected ovary is the hallmark of
ovarian torsion. A variety of technical pitfalls, however, may result in lack of 4ow,
including inadequate depth of penetration, improper Doppler gain, and inappropriately
elevated pulse repetition frequency. Conversely, 4ow 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 4ow may still be present, although usually
asymmetrically decreased compared with the contralateral ovary. Lack of venous 4ow in
the symptomatic ovary is a more useful diagnostic indicator than the absence of arterial
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 signi cantly enlarged and relatively solid appearing with multiple small
peripheral follicles and normal Doppler 4ow (Figure 2-18). Some of these cases are
caused by subtotal torsion, and if blood 4ow is present, the ovary may be successfully
untwisted at surgery. Some cases seem to be caused by a hemorrhagic corpus luteum.
15Usually patients are treated symptomatically.!
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
3follicles measuring 2 to 9 mm in diameter or increased ovarian volume (>10 cm ).
Although increased stroma and increased stromal echogenicity are common ndings,
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
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, calci cation, 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- lled
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
17occasionally a lipoleiomyoma (an uncommon fat-containing myoma) (Figure 2-22).
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
calci cations, hemosiderin, or specular re4ections from tiny cysts below the spatial
resolution of the transducer (Figure 2-23). Slightly larger focal calci cations, usually 1
cm or smaller, are also common. Retrospective studies have not shown interval
18development of neoplasms.
FIGURE 2-23 Echogenic foci in a postmenopausal ovary. Transvaginal ultrasound
image reveals tiny benign echogenic foci (arrows) related to specular re4ections 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 broma–thecoma tumors (Figure 2-24). These
tumors are composed of brous cells with varying amounts of thecal cells and thus are
often similar in appearance to uterine myomata. We have found two helpful maneuvers
to di erentiate 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 con rm 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 broid 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
realtime scanning to derive the greatest bene t 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 ofpenetrating, 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
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
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.
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 TomographyChapter 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
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 di) erent
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 signi1cantly 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
di4 cult to clearly distinguish the uterus from the cervix, the two can be separated from
one another by their con1gurations because the uterus is somewhat triangular in shape,
whereas the cervix has a more rounded appearance. The cervix enhances to a lesserdegree 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
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 : uid (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
2is better at detecting gross thickening. Additionally, review of CT often leads to a
falsepositive conclusion for true endometrial thickening; therefore transvaginal sonography
2should always be used to con1rm the CT 1ndings. 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 (Figure3-5, A) and can be misinterpreted as thickened. When such a 1nding is seen, sagittal
reconstruction images should be interpreted in addition to the axial images because they
2can typically confirm the absence of true endometrial thickening (see Figure 3-5, B).
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
On postcontrast images the uterus can have di) erent 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 becominghomogeneous on more delayed imaging. Zonal patterns of enhancement that have been
described include subendometrial (Figure 3-6), peripheral myometrial (Figure 3-7), and
3faint di) use myometrial enhancement (see Figure 3-2). 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 o) multiple branches to the uterus as it courses between the layers of broadligament, 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 speci1city in characterization of müllerian duct anomalies.
Occasionally the diagnosis can be made on CT images, usually seen as incidental 1nding
(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 1nding 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 protrudesinto 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 1brous tissue. These should be
di) erentiated 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
3 3volume is 9.8 cm ; in postmenopausal women, 5.8 cm ; and in the premenarchal group,
3 43.0 cm . In menstruating women, the normal ovary can be identi1ed on CT in most
instances (Figure 3-14). Postmenopausal ovaries may be di4 cult 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
1Figure 3-14). 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-desac, pelvic inlet, iliac fossa, and lower abdomen (Figure 3-16). During the 1rstpregnancy, 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
27year-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
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
22year-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 1nding 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,
5crenulated, and hyperenhancing wall (see Figures 3-1 and 3-17). 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 athick, 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 di) erentiating between ovarian and nonovarian masses (Figure 3-18). The
6presence of this finding has been described as the ovarian vascular pedicle sign.
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), con1rming the
ovarian origin of the mass.
Ovarian Vein Reflux
Re: ux of contrast material from the left renal vein into the left ovarian vein is commonly
7seen in asymptomatic patients and can be present in up to 50% of multiparous women.
Depending on the grade of re: ux, retrograde opaci1cation can reach the parauterine
veins, and some patients can develop parauterine varices (Figure 3-19). Although the
re: ux 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 re: ux. Computed tomographic image of a 39-year-old
multiparous woman shows re: ux of contrast material into the dilated parauterine veins
The vagina surrounds the inferior portion of the cervix creating anterior, posterior, and
lateral fornices. These can be identi1ed 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 : uid, obstruction and 1stula should
be considered. On postcontrast imaging the vaginal mucosa enhances intensely (Figure
320) 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).
LigamentsSeveral fascial condensations referred to as ligaments support the female genital organs
1and can be seen on CT examinations (Figure 3-21).
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 1mbria) 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.
T h e round ligament (Figure 3-23) consists primarily of 1bromuscular 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 theinguinal 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-o) 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 1nding. 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 1bers fuse medially with the posterior 1bers
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
The ovarian ligament (Figure 3-26) and suspensory ligament of the ovary (see Figure
318) 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 signi1cance 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 1rst site of disease extension (Figure
328). 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
8diameter. 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 speci1city but
have more false-negative results. This obstacle has been somewhat overcome by adding
positron emission tomography (PET) imaging to CT, which increases speci1city and
sensitivity for detection of malignant lymph nodes.
Posthysterectomy Patient
CT is commonly performed in women who have undergone previous hysterectomy. A CTexamination 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 cu) 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 di4 cult 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 1nding
and can occur in up to 80% of cases in postoperative patients if hysterectomy and
9oophorectomy have been performed for malignant disease (Figure 3-30). 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 cu) . 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 1lling 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 1eld, most
commonly lateral to the paracolic gutter or lateral to the colon in the iliac fossa (Figure
331, 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. Identi1cation 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 : udeoxyglucose
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 1nding 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 di) erential 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
10thinner slices (3 mm) and multiplanar images are used (Figure 3-32).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
11staging, but MRI surpasses CT in tumor visualization and local extension.
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 in1ltration 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
12local extent.
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 in: ammatory 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
1rst, unless bene1ts 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 therisk 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 1rst 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
13it is likely not necessary to disrupt the breast-feeding process after the examination.
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, speci1c precautions should be taken to minimize the
Technique Description
Adequate opaci1cation 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 opaci1cation 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
e) ect 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 in: ammatory 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 de1ne pelvic structures as a result
of patterns of parenchymal opaci1cation and helps di) erentiate 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 di) erentiate pelvic lymph nodes from the vessels because pelvic vessels usuallyopacify 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 o) ers 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 unopaci1ed 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 opaci1cation of the bowel. Similarly,
adequate opaci1cation of small bowel is important to di) erentiate 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 de1ne 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 unopaci1ed loop of bowel in the absence of oral
contrast. B, Presence of oral contrast con1rms 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
FIGURE 3-35 Pelvic in: ammatory 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 1rst 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
14spatial resolution owing to submillimeter section collimation. Section collimation is
determined by available detector con1guration. 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
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
15reformatted data sets of high quality.
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 a) ect 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 su4 ce. 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
milliamperesecond settings to overcome the increase in image noise that occurs at lower kilovolt peak
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 fordisplaying 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 in1ltrating 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 a) ect the radiation dose to the patient. One of the
most e) ective 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.KEY POINTS
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 gadoliniumcontrast 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.
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.