MRI of the Female Pelvis

MRI of the Female Pelvis

1. 1115
MRI of the Female Pelvis: A
Review
Hedvig Hricak1 MRI of the female pelvis offers a unique display of pelvic anatomy. On the T2-weighted
images, uterine zonal anatomy and temporal changes under hormonal stimuli are well
displayed. Benign and malignant uterine neoplasms are accurately demonstrated. How-
ever, tumor type cannot be diagnosed. In the staging of endometrial and cervical
carcinoma, MRI offers distinct advantages over sonography and/or CT. The main
advantages are the abilities to depict tumor and to provide great tissue contrast between
the tumor and surrounding normal tissue. In the evaluation of ovarian tumors, experience
with MRI is still limited. This is only the beginning of the clinical applications of MRI.
Much more work needs to be done to explore fully the value of this versatile and
powerful technique.
In the short time since MRI has become a clinical procedure, it has shown
advantages over existent diagnostic techniques. In the imaging of the female pelvis,
MRI has complemented sonography and/or CT in further refining anatomic details
and allowing the display of disease in more detail. Sonography remains the
screening technique for the many uterine and ovarian afflictions. However, its
significant operator dependence, limitations due to patient habitus, and relative
inability to determine tissue characterization, significantly decrease its value [1-3].
CT, which is currently used for the staging of pelvic neoplasms, also has limitations.
These include the presence of ionizing radiation and distortion due to metallic clips,
contrast media, and bone density. All this restricts the usefulness of pelvic CT,
particularly in regard to soft-tissue resolution [4]. The advantages of MRI in the
study of pelvic abnormalities have recently been described [5-12]. MRI is nonin-
vasive, does not depend on ionizing radiation, has superb soft-tissue contrast
resolution, and is capable of multidirectional imaging, simultaneous imaging of
multiple sections, and visualization of blood vessels without the need for contrast
injection. Images of the pelvis, as compared with those of the abdomen, are better
quality because respiratory and intestinal motion are at a minimum. However, MRI
also has certain limitations. These include a relatively long scanning time and the
contraindication to scan patients with cardiac pacemakers, intracranial vascular
clips, and large metallic devices.
Only when the advantages and limitations of MR are understood can the
technique be fully used.
Normal Anatomy
Received February 11, 1986; accepted February
14, 1986. The ability to optimally discern anatomic details depends on instrument param-
‘Department of Radiology, University of Califor- eters and the plane of section used. With the 0.35-T MT/S Diasonics (Milpitas,
nia Schoo! of Medicine, University of California at CA), images are obtained using spin-echo sequences with repetition time (TA) of
San Francisco, San Francisco, CA 94143. . .
0.5 sec and echo-delay time (TE) of 30 msec. With this technique, the uterus is
1986 imaged as a homogeneous, medium-signal-intensity structure, and the uterine
© American Roentgen Ray Society zonal anatomy is indistinct (Fig. 1). On the T2-weighted sequence (TA = 2 sec. TE Downloaded from www.ajronline.org by 113.161.77.46 on 11/13/18 from IP address 113.161.77.46. Copyright ARRS. For personal use only; all rights reserved

2. 1116
HAICAK
AJR:146, June 1986
Fig. 1 -Normal uterus in woman of reproductive age. A, TR = 2 sec, TE =
60 msec. B, TR = 0.5 sec, TE = 30 msec. Corpus uterus is seen upon dome
of empty urinary bladder (B). Myometrium (m) can be clearly differentiated from
endometrium (e) on T2-weighted image (A). Arrow shows junctionai zone.’
On Ti -weighted image (B), zonal anatomy is obscured.
= 60 msec), the uterus can be divided into corpus, isthmus,
and cervix 14-1 0J. Within the corpus, the myometrium
and endometrium are imaged with high signal intensity sepa-
rated by a “junctional zone,” a low-intensity line between
them. The “junctional zone” is believed to reflect the vascular
structures, mainly veins, located within the inner third of the
myometrium [8-1 01. The MRI appearance of the corpus
uterus is markedly influenced by hormonal stimuli [10]. In
women of reproductive age, the appearance of the uterus
changes during the menstrual cycle [1 0]. The endometnum
and myometrium, separated by the “junctional zone,” are
always seen as distinct zones. However, the endometrium
changes in width and is widest in the midsecretory phase.
The volume and signal intensity of the myometrium changes
also. On the T2-weighted image, the signal intensity of the
myometrium is higher during the secretory phase. Also, the
total uterine volume changes during the menstrual cycle and
is the greatest during the secretory phase. Women of repro-
ductive age taking oral contraceptive pills have a different
MAI appearance of the uterus. In these women, the myomet-
rium and endometrium separation is indistinct [10] (Fig. 2).
Also, endometrial atrophy is marked, and the junctional zone
is not consistently seen. Images of the premenarchal and
postmenopausal uterus differ from reproductive-age uteri.
Premenarchal and postmenarchal uteri have a small corpus
and atrophic or absent cycling endometrium; the length of the
corpus equals that of the cervix [5-10] (Fig. 3). The uteri of
postmenopausal females taking exogenous estrogen have an
MA appearance similar to the uteri of women of reproductive
age.
The length and orientation of the normal cervix vary. The
best depiction of the cervix is on the T2-weighted image. The
normal cervix has two separate zones. A central zone imaged
with high signal intensity represents the cervical epithelium
Fig. 2.-Normal volunteer of reproductive age taking oral con-
traceptive pills. TR = 2 sec, TE = 56 msec. Corpus uterus is
globular in configuration. Myometrium (m) is of high signal intensity,
and differentiation between myometnum and endometrium is not
possible. Two small leiomyomas (arrows) are identified. Normal-
appearing cervix (c) with distinction between cervical stroma and
central mucus.
Fig. 3.-Postmenopausal uterus of normal volunteer. TR = 2 sec.
TE = 40 msec. Corpus length similar to cervix length (c). Within corpus,
differentiation between myometnum (m) and endometnum (arrow) is
poor. Myometnum is of lower signal intensity than the myometnum in
women of reproductive age. This normal volunteer was 10 years
postmenopausal.
and mucus. It is surrounded by a cylinder of low-intensity,
fibrous cervical stroma [5] (Figs. 2 and 4). The parametrium
is imaged with medium-high signal intensity and is easily
distinguished from the low-intensity cervical stroma [5, 11] Downloaded from www.ajronline.org by 113.161.77.46 on 11/13/18 from IP address 113.161.77.46. Copyright ARRS. For personal use only; all rights reserved

3. Fig. 4.-Antefiexed, horizontally positioned cervix. On sagittal image. TA =
2 sec, TE = 60 msec. Honzontal orientation of cervix (C) is seen. Cervical
stroma imaged wrth low signal intensity can be separated from higher intensity
mucus. V = vagina; B = urinary bladder. On the transverse image (B) (TA =
0.5 sec. TE = 30 msec), intensity of cervix (c) blends with surrounding
parametrium (p). Uterine vessels = arrowhead. Transverse image (C) (TA = 2
sec, TE =60 msec), Obtained at same anatomic location as 4B; cervical stroma
is clearly separated from surrounding parametna, which are highly vascular and
show increased signal intensity on this long-TA second-echo image.
(black arrowhead). Levator ani (open arrows). B, in the same patient, a section
2 cm more cephalad shows vaginal fomices (arrowheads). This demarcates
upper third of vagina. Urinary bladder = B.
AJR:146, June 1986 MRI OF FEMALE PELVIS
1117
Fig. 5.-Normal vagina. A, TA = 2 sec, TE = 60 msec. Anatomic location
of lower third of vagina (black arrowhead) marked by anterior-placed urethra
(black arrow). tkethra is clearly separated from posterior low-intensity vagina
(Fig. 4). On the second-echo image (TA = 2.0 sec, TE = 60
msec), a high signal intensity from slow-flowing blood is often
detected within the parametrium. Although a longitudinal cer-
vical os is most common in the sagittal plane (Fig. 4), some-
times the long axis is in the horizontal plane. Either plane of
cervical orientation is easily depicted by MRI.
The vagina can be identified separately from the surround-
ing structures on the T2-weighted image (Fig. 5). It is imaged
with a high-intensity center, representing the vaginal epithe-
hum and mucus, and a lower-intensity wall. The anatomic
division between the lower and upper thirds of the vagina is
easily seen on transverse images. The lower third of the
vagina corresponds to a plane of section below the base of
the bladder (Fig. 5A). Its anatomic level is marked by the
anteriorly placed urethra. The middle third of the vagina
corresponds to the level of the bladder base. The upper third
of the vagina is demarcated by the lateral vaginal fomices
(Fig. SB).
The normal ovaries are more difficult to demonstrate on
MRI. They have a low to medium signal intensity on the Ti -
weighted image (short TR and TE) [12] (Fig. 6). When TA is
short (0.5 sec), distinguishing ovaries from the surrounding Downloaded from www.ajronline.org by 113.161.77.46 on 11/13/18 from IP address 113.161.77.46. Copyright ARRS. For personal use only; all rights reserved

4. Fig. 6.-Normal ovaries. A, On Ti-weighted image (TA = 0.5, TE = 30
msec), ovaries have medium signal intensity, and distinguishing ovary from
surrounding bowel loops or uterus (U) is difficult. Aight ovary (0). Laterally
positioned ovarian vessels (arrow) serve as landmark. Urinary bladder = B. B,
With prolongation of TA at 2 sec and second echo image (TE = 60 msec),
signal intensity of ovaries increases and approaches that of surrounding fat.
Ovarian vessels (arrow) are identified as dark tubuiar structures at periphery
of adnexa.
1118 HRICAK
AJA:146, June 1986
bowel loops is difficult. When TA is longer, ovarian signal
intensity increases and approaches that of the surrounding
fat. The sagittal plane is ideal for demonstrating the uterus,
but not the ovaries. The anatomic landmarks, the vessels
surrounding the ovaries identified as dark tubular structures
at their periphery, are easily seen in coronal or transverse
planes. Although there are difficulties in imaging the ovaries,
normal ovaries are demonstrated on MR in 87% of women of
reproductive age when continuous slices without gap are
used and when the plane of imaging is either coronal or
transverse [12].
Pathology
Uterine Leiomyoma
Leiomyoma is the most common uterine tumor, occurring
in 20 to 30% of women during their reproductive years [13].
These tumors may be solitary or multiple and are found in
submucosal, intramural, or subserosal sites of the uterine
corpus or cervix. In the evaluation of leiomyoma, sonography
often augments the pelvic examination. However, the sono-
gram may appear normal in 22% of the cases of leiomyoma
[1]. The examination is of limited value when tumors are
small, when the uterus is retroverted or retrodisplaced, and
when there is coexisting ovarian disease [1-3]. MRI can
provide more accurate assessment of the number, size, and
precise location of Ieiomyomas. This is important in the clinical
settings of infertility or recurrent abortion, or before myomec-
tomy [14]. Tumors as small as 0.5 cm are accurately dem-
onstrated by MAI (Fig. 7), and the precise location of the
mass in either a submucosal, myometnal, or subserosal p0-
sition is clearly displayed (Fig. 8).
The optimal imaging sequence for the diagnosis of an
intramural and submucosal Ieiomyoma is the T2-weighted
image, which renders the best contrast between the tumor
and the myometnum or endometnum (Figs. 7 and 8). A
combination of Ti - and T2-weighted sequences is always
desirable. However, both Ti - and T2-weighted sequences
are required in the evaluation of subserosal lelomyomas. The
Ti -weighted image will make tumor distinction from adjacent
adipose tissue optimal, while the T2-weighted image is
needed for the assessment of tumor architecture and distinc-
tion of tumor from normal myometnum.
In correlating MRI features of leiomyoma with histologic
characteristics, two main groups of Ieiomyomas can be iden-
tified: degenerative and nondegenerative lelomyomas (Figs.
2, 7, and 8). Nondegenerative leiomyomas have uniform echo
distribution, and their signal intensity is similar to or slightly
lower than that of myometnum on the Ti -weighted image.
On the T2-weighted image, the signal intensity is considerably
lower than that of adjacent myometnum or endometnum.
Degenerative leiomyomas demonstrate a spectrum of signal
intensities ranging from low to high. Degenerative leiomyomas
usually have a heterogeneous signal intensity on the T2-
weighted image. The type of degeneration (either hyaline,
myxomatous, or fatty) cannot be differentiated by MRI. Fur-
thermore, MRI cannot distinguish between benign and malig-
nant tumor degeneration [14].
In view of the modem surgical approach to leiomyoma, the
accuracy of MRI in the diagnosis of submucosal tumors,
which are a known cause of excessive uterine bleeding,
infertility, and abortion, can have a significant clinical impact.
Noninvasive identification of small lesions within or adjacent
to the endometnal cavity can guide the surgical approach
through the hysteroscope in circumstances in which uterine
conservation is desired. The MRI study can obviate multiple
dilatation-and-curettages and/or hysteroscopy. Downloaded from www.ajronline.org by 113.161.77.46 on 11/13/18 from IP address 113.161.77.46. Copyright ARRS. For personal use only; all rights reserved

5. lultiple leiomyomas. 1 . = 2.0 sec. TE = LB msec. Various-size
leiomyomas (the largest one (L) seen anteriorly) are well visualized adjacent to
higher-intensity myometrium because of their low signal intensity. Small ieiom-
yoma (0.3 cm) (arrow) is well seen. Fundal leiomyoma (open arrow) is calcified
and creates no signal. “Junctional zone” is intact and endometrium (e) is well
displayed. Normal cervix (c). Reprinted with permission from Hricak et al. (8].
. . lelomyomas. Uterus = U. Smaller anterior leiomyoma
(L) and larger posterior subserosal leiomyoma (LL). Within larger posterior
subserosal leiomyoma, a central area of high signal intensity (curved arrow)
represents area of cystic degeneration. Reprinted with permission from Hricak
et al. (14].
Fig. 9.-Endometrial carcinoma in a postmenopausal patient (68 years old);
not taking exogenous hormones. A, Sagittal image, TA = 1.0 sec. TE = 28
msec. Endometrial cavity is expanded and filled with medium-signal-intensity
tumor (T). Myometrium (arrow) is compressed and imaged with low signal
AJA:146, June 1986 MRI OF FEMALE PELVIS 1119
Endometrial Carcinoma
Carcinoma of the endometrium is the most common inva-
sive carcinoma of the female genital system and the fourth
most frequent malignancy in American women [15]. Therapy
intensity. Tumor extends into cervix (C). Bladder = B. B, TA = 1 sec. TE = 56
msec. On second-echo image, the contrast between myometrium and tumor is
enhanced. Margins of myometrium are intact, indicating that there is no
transmyometrial tumor extension.
for endometrial carcinoma is determined by its stage; thus, a
reliable method for evaluation of the presence and extent of
carcinoma is essential [16]. The staging system of the Inter-
national Federation of Gynecology and Obstetrics (FIGO) is a
clinical system and is suboptimal compared with surgical Downloaded from www.ajronline.org by 113.161.77.46 on 11/13/18 from IP address 113.161.77.46. Copyright ARRS. For personal use only; all rights reserved

6. Fig. 1 0.-Endometrial carcinoma. Transverse image, TA = 2 sec. TE = 60
msec. On this T2-weighted image, uterus is enlarged for patient’s age. There
is tumor (T) in corpus. Surrounding myometrium (m) is intact and no transmy-
ometrial tumor invasion is detected. Ascites = A. Simple ovarian cyst (Cy).
Fig. 11 -Cervical carcinoma (TA = 2 sec. TE = 60 msec). Tumor (T) is
expanding cervical canal and extending to upper vagina. Aemaining cervical
stroma (arrowhead), imaged with low signal intensity, is intact. Wnary bladder
= B.
1120
HRICAK AJA:146, June 1986
staging. Radiologic techniques, such as CT and sonography,
are unreliable for staging endometrial carcinoma [1 7]. On
MRI, endometnal carcinoma is detected as an abnormality in
the central endometrial cavity. Widening of the endometrium
or the presence of endometrial masses are important findings
[1 8] (Figs. 9 and 1 0). In postmenopausal women not taking
exogenous hormones, high signal intensity in the central canal
can be seen only on the T2-weighted image and is never
greater than 2 mm in width. In endometnal carcinoma patients,
the central high-intensity endometnum is wider. The disruption
of the low-intensity line (“junctional zone”) between the myo-
metrium and endometnum may be an important indication of
myometnal invasion. The junctional zone is always seen in
normal premenopausal women, but can be absent in normal
postmenopausal women. Preservation of the junctional zone
is an excellent indicator that endometnal carcinoma is con-
fined to the endometrium. The junctional zone is absent in
every patient in whom myometnal invasion is found. Segmen-
tal nonvisuahization of the junctional zone is also a reliable
finding of transmyometrial invasion. However, as complete
nonvisualization of the junctional zone can occur with normal
postmenopausal women, nonvisualization of the junctional
zone should be coupled with additional findings before the
diagnosis of transmyometrial invasion is made [18]. MRI
cannot differentiate between endometnal carcinoma and ad-
enomatous hyperplasia. Histologic diagnosis is required.
When endometnal carcinoma is histologically documented,
the local staging of the disease by MRI is excellent and the
uterine size and invasion of the myometrium or cervix can be
clearly demonstrated [8, 18]. The MRI depiction of lymphad-
enopathy is similar to that of CT [19, 20]. Both techniques
rely on size of the node, and neither MRI nor CT can differ-
entiate malignant from hyperplastic node enlargement [21].
Carcinoma of the Cervix
Invasive carcinoma of the uterine cervix is the most com-
mon malignancy of the reproductive tract in women under the
age of 50 with a yearly incidence of 1 6,000 new cases and
7400 deaths [22]. Clinical examination and currently used
radiologic studies, including sonography and CT, are limited
in providing accurate assessment of tumor extent [1 6]. The
stage of the disease, with emphasis on whether the tumor is
confined to the cervix or extends to the parametna, is the
most critical factor in determining the optimal therapy and the
prognosis [23].
MRI has excellent sensitivity in depicting the neoplasm and
separating tumor from cervical stroma. The soft-tissue dis-
crimination enhances the MRI accuracy in the staging of
localized cervical neoplasm [1 i]. Once cancer is demon-
strated, MAI is useful in determining tumor extent, including
involvement of the cervical stroma, vagina, or parametnum
(Fig. ii). On the T2-weighted image, the cervical neoplasm
has a high signal intensity clearly distinguishable from that of
the cervical surrounding normal tissue. On the Ti -weighted
image, the cervical mass is isointense with the normal cervix,
and only gross parametnal or ovarian extension causing
contour abnormalities can be visualized. Vaginal involvement
must always be assessed in two planes perpendicular to each
other. It is detected as a mass interrupting the normally low-
intensity vaginal wall. Parametnal extension is diagnosed by
either an asymmetric appearance of the parametnum or by
abnormal tumor intensity extending into the parametnal re-
gion. Downloaded from www.ajronline.org by 113.161.77.46 on 11/13/18 from IP address 113.161.77.46. Copyright ARRS. For personal use only; all rights reserved

7. AJA:146, June 1986 MRI OF FEMALE PELVIS
1121
Fig. 12.-Patient with multiple congenital anom-
aiies. Normal right uterus. Myometrium (m) is clearly
separated from central high-intensity endometrium
(e). Normal cervix (C). Two masses are seen in the
midline. Lower one presents second uterus (u) dis-
tended with blood. Cervix (asterisk) of second
uterus is fibrotic, and no cervical canal can be seen.
Large high-signal-intensity mass (E) is endome-
trioma (surgicaily proved).
Ovarian Disease
The direct multiplanar imaging capability of MRI is particu-
larly useful in determining the ovarian or uterine origin of a
pelvic mass. When a mass is very large, determination of its
origin can be difficult. The ability of MRI to see consistently
the uterine zonal anatomy facilitates depiction of the origin of
the pelvic mass (Fig. 1 2). The transverse or coronal plane or
a combination of the two planes appears to be the most
useful in evaluating ovarian disease.
Simple ovarian cysts appear as well-circumscribed homo-
geneous masses with a smooth interface (Fig. 1 3). When
peripherally located, simple cysts have smooth and almost
imperceptible walls. All simple ovarian cysts are imaged with
a low signal intensity on the Ti -weighted image. Their signal
intensities increase on the T2-weighted images. An ovarian
hemorrhagic cyst is seen as a well-circumscribed homoge-
neous mass with a smooth wall of varied thickness (Fig. 12).
The signal intensity of hemorrhagic ovarian cysts varies with
their age. Acute hemorrhagic cysts have an intermediate
signal intensity on Ti -weighted images and a high signal on
T2-weighted images. Chronic hemorrhagic cyst shows high
signal intensity on the short-TA, short-TE (Ti-weighted) im-
age. Its signal intensity is similar to that of fat on the T2-
weighted image (Fig. 12).
MRI can unequivocally distinguish between simple fluid (no
protein content, no hemorrhage) and all other types of lesions.
However, the presence of simple or hemorrhagic fluid does
not specify a particular gynecologic process, and MRI cannot
differentiate benign from malignant ovarian lesions. Further-
more, with current techniques, MRI cannot consistently dif-
ferentiate old hemorrhagic fluid from fat. This is important,
since some dermoid cysts consisting of fat produce a high
signal intensity regardless of which TA or TE is used and
have been reported to be indistinguishable from fat [12]. MAI
Fig. 1 3.-Simple ovarian cyst (TA = 2 sec. TE = 30 msec). Patient has
massive ascites (A). Ovarian cyst (Cy) has homogeneous low signal intensity
suggestive of simple fluid content. Findings were proven at surgery.
cannot consistently distinguish between the solid component
of hemorrhage and malignant lesions.
Conclusion
MAI is emerging as an important imaging technique in the
study of the female pelvis. Sonography remains the best
screening procedure for the evaluation of a suspected pelvic
mass. Precise localization and tissue differentiation, however,
are better achieved with MAI. Therefore, when a sonogram
is suboptimal, the origin of a pelvic mass is not established,
or when differentiation between a simple-fluid lesion and
another type of ovarian tumor requires further clarification,
MRI can be useful. Currently, the most important role of MRI
is in the staging of endometnal and cervical neoplasms. The
role of MRI in the study of ovarian tumors needs further
investigation. While many applications of MRI are already
established, clinical experience is still limited, and the full
potential of MRI for the evaluation of the diseases of the
female pelvis has not yet been determined.
REFERENCES
1. Gross BH, Silver TM, Jaffe MH. Sonographic features of uterine
leiomyomas. J Ultrasound Med 1983;2(Sept):401-406
2. O’Brien WF, Buck DR, Nash JD. Evaluation of sonography in the
initial assessment of the gynecologic patient. Gynecol Obstet
Invest 1984;149:598-602
3. Walsh JW, Taylor KJW, Wasson McI, et al. Gray-scale ultrasound
in 204 proved gynecologic masses: accuracy and specific diag-
nostic criteria. Radiology 1979;1 30:391-397
4. Lee JKT, BaleD. Pelvis. In Lee JKT, Sagel SS, Stanley RJ, eds.
Computed body tomography. New York: Raven Press,
1983:393-419 Downloaded from www.ajronline.org by 113.161.77.46 on 11/13/18 from IP address 113.161.77.46. Copyright ARRS. For personal use only; all rights reserved

8. ii 22 HAICAK AJA:146, June 1986
5. Hncak H, Alpers C, Crooks LE, Sheldon PE. Magnetic resonance
imaging of the female pelvis: initial experience. AJR 1983;
141:1119-1128
6. Bryan PJ, Butter HE, LiPuma JP, et al. NMR scanning of the
pelvis: initial experience with a 0.3T system. AJR 1983;
141 :1111-1118
7. Butler H, Bryan PJ, LiPuma JP, et al. Magnetic resonance
imaging of the abnormal female pelvis. AJR i984;1 43:129-
1266
8. Hncak H, Schriock E, Lacey C, et aI. Gynecologic masses-
value of MRI. Am J Obstet Gynecol 1985;1 53(1):31 -37
9. Lee JKT, Gersell DJ, Balfe DM, Worthington JL, Picus 0, Gapp
G. The uterus: in vitro MR anatomic correlation of normal and
abnormal specimens. Radiology i985;1 57:175-179
10. Demas B, Hricak H, Jaffe RB. Uterine MR imaging: effects of
hormonal stimulation. Radiology 1986;1 59:123-126
11. Winkler ML, Hncak H. Magnetic resonance evaluation of cervical
carcinoma. Presented at the annual meeting of the Radiological
Society of North America, Chicago, November 1985
12. Dooms GC, Hricak H, Tscholakoff D. Magnetic resonance im-
aging of adnexal structures: normal and pathologic. Radiology
1986;1 58:639-646
13. Silverberg SH, ed. Principles and practice of surgical pathology.
New York: Wiley & Sons, 1983:1323-1327
14. Hricak H, Tscholakoff D, Heinrichs L, et al. Uterine leiomyoma-
correlation by magnetic resonance imaging, clinical symptoms
and histopathology. Radiology 1986;1 58:385-391
1 5. Boronow RC, Morrow CP, Creasman WT, et al. Surgical staging
in endometrial cancer: clinical-pathologic findings of a prospec-
tive study. Obstet Gynec 1984;63(6):825-832
16. Berman ML, Ballon SC, Lagasse LD, Watson WG. Prognosis
and treatment of endometnal cancer. Am J Obstet Gynecol
1985;1 36(5):679-688
17. Kerr-Wilson RM, Shingleton HM, Orr JN. The use of US and CT
scanning in the management of the gynecologic cancer patient.
Gynecol Oncol 1984;18:54-61
18. Hncak H, Fisher MR. Shapeero LG, et al. MRI in the evaluation
of endometrial carcinoma and its staging. Presented at the annual
meeting of the Radiological Society of North America, Chicago,
November 1985
19. Dooms GC, Hricak H, Crooks LE, Higgins CB. Magnetic reso-
nance imaging of the lymph nodes: comparison with CT. Radiol-
ogy 1984;153:719-728
20. Lee JKT, Heiken JP, Ling 0, et al. Magnetic resonance imaging
of abdominal and pelvic lymphadenopathy. Radiology 1984;
153:181-188
21. Dooms GC, Hncak H, Moseley ME, Bottles K, Fisher M, Higgins
GB: Characterization of Iymphadenopathy by magnetic reso-
nance relaxation times: preliminary results. Radiology 1985;
155:691-697
22. Silverberg E. Cancer statistics, 1980. CA 1980;30:23-38
23. Ferenczy A. Carcinoma and other malignant tumors of the cervix.
In: Blaustein A, ed. Pathology of the female genital tract, 2nd ed.
New York: Springer-Verlag, 1982:184-222Downloaded from www.ajronline.org by 113.161.77.46 on 11/13/18 from IP address 113.161.77.46. Copyright ARRS. For personal use only; all rights reserved