Ruben Richard

Kier, MD #{149} Shirley R. Viscarello, MD

Pelvic

M. McCarthy, MD, E. Schwartz,

Masses

terms:

Endometniosis,

ry, cysts, 852.311 Pelvis, MR studies, plications,

85.1214

852.3192

Ovary, neoplasms, 85.1214 #{149} Pregnancy, 85.3 #{149} Pregnancy, MR studies, #{149} Teratoma, 852.313

Radiology

1990;

#{149}

Ova852.313 com#{149}

176:709-713

From the Departments of Diagnostic Radiology (R.K., 5MM., L.M.S.) and Obstetrics and Gynecology (R.R.V., P.E.S.), Yale University School of Medicine, 333 Cedar St. P0 Box 3333, New Haven, CT 06510. Received February 13, 1990; revision requested April 16; revision received May 9; accepted May 10. Address reI

print C

requests RSNA,

to R.K. 1990

#{149} Leslie

M. Scoutt,

in Pregnancy:

The value of magnetic resonance (MR) imaging was assessed for 17 pregnant patients with sonograms suggestive of a pelvic mass. The MR imaging signal features improved lesion characterization in 47% (eight of 17) of cases, including two of four mature cystic teratomas of the ova17, three uterine fibroids, one solid ovarian tumor, one endometnioma, and a distended urinary bladder that had been mistaken for an ovarian cystic mass. Both MR imaging and sonography were accurate for the characterization of three ovarian cystadenomas and two simple ovarian cysts. On both MR images and sonograms, two simple ovarian cysts were incorrectly diagnosed as complex cystic masses and one teratoma was incorrectly diagnosed as a simple cyst. The origin of the pelvic mass (13 in the ovary, three in the uterus, and one distended urinary bladder) was accurately determined on 100% (17 of 17) of the MR images versus 71% (12 of 17) of the sonograms. In three cases, the results of MR imaging led to cancellation of surgery, which would have proceeded on the basis of the sonographic results alone. MR imaging is a valuable complement to sonography for preoperative evaluation of pelvic masses in pregnant patients. Index

PhD MD

#{149} Peter

O

MR

neoplasms

VARIAN

0.1%-0.4% During

the

common

occur

first

pelvic

mass

the

is the

corpus

Beyond

trimester,

the

first

plasms,

are

most

pregnancy,

are malignant can

Physical

examination the

first

even

com-

of the

trimester

pelvis

is difficult.

Therefore, the organ of origin of a pelvic mass or its cystic on solid natune

may

be indeterminate.

dominal

sonography

tinguish

cystic

characterization the gravid uterus mass. Computed not a desirable nant

patients,

radiation masses

and have

help

solid

dis-

masses,

but

since

CT

since

most

(CT) pneg-

for

uses

is

studies

obstetrics signed imaging masses

have

appear-

(12-17).

explored

Our

its

study

role

was

AND

pregnancy,

either

MR

such

cysts

imag-

are

fibroids at sonografor MR imaging,

usually

is not

performed

for fibroids. during the

and

imaging

in

Four

scanning

MR imagsecond or

gynecology,

high-risk

patients

of radiology,

in

de-

exami-

nation or at sonography. During the penod of the study, all 15 pregnant patients considered for operative intervention for a suspected pelvic mass by the department of obstetrics at our institution were referred for MR imaging. Two additional pregnant patients with pelvic masses

obstet-

underwent

transabdominally

ment

in

the

including

depart-

one

patient

also underwent scanning in the periunit. In two patients who underscanning at other institutions, the sonographic

tenpretation In both

METHODS

at physical

with

images

viewed by an experienced who was blinded to both

Seventeen pregnant women, aged 1636 years, were referred for MR imaging to evaluate a pelvic mass seen on a sonogram. These masses were first noted duning

that

at sonog-

up with serial sonogsurgery. Similarly, pathat were confidently

of obstetrics

original

to assess the usefulness of MR in the evaluation of pelvic in pregnant patients.

MATERIALS

surgery

cyst

with use of both transabdominal and endovaginal imaging. These scans were obtamed as part of a prospective study of

who natal went

ionizing

ance on CT scans (5). Magnetic resonance (MR) imaging has proved effective in defining many gynecologic disorders (5-11), and some preliminary

with

as uterine not referred

tic patients.

pelvic

a nonspecific

studied

routinely followed raphy rather than tients with masses

endovaginal

is often limited, as may conceal the tomography alternative

not

a simple

patients

partment

Transab-

may

from

were

obstetricians.

6 cm in diameter

third trimester of pregnancy; patients referred during the first trimester were scheduled for MR imaging after the 15th week of pregnancy. Sonography was performed on all 17 patients. Twelve patients underwent scanning at the perinatal unit of the de-

be-

obstetric

for

during pregnancy ing was performed

of these

and

cause

criteria

since

since

during

2%-5%

ing,

outside

than

the

raphy

by

smaller

diagnosed phy were

cys(1-3).

performed

because

nign masses plications. beyond

neo-

benign

or cystadenomas

is usually

masses

major-

ovarian

commonly

tic tenatomas

Surgery

by (4).

the

cystic

lu-

referred

Masses met

most

usually regresses week of gestation

of masses

were

(1-3).

trimester,

.

Imaging’

in

of pregnancies

teum cyst, which the 10th to 15th ity

MD

were

and the operative cases, this interpretation

imaging

GE Medical body

coil

Systems, was

report on a 1.5-T system

(Signa;

Milwaukee).

used

for

in-

findings. agreed

with the original sonographic from the other facility. MR imaging was performed superconducting

re-

radiologist the original

both

The

excitation

and signal reception. T2-weighted spinecho imaging was performed in the axial plane with a repetition time of 2,000 msec and an echo time of 80 msec and in the sagittal plane with a repetition time of 1,700

msec

and

an

All T2-weighted echoes, with the with

an

echo

patients

also

time

patients),

patient), msec

with and

an

of 80 msec.

employed performed

or the time

two

Thirteen

Ti-weighted

in either the the coronal

a repetition echo

time

of 20 msec.

underwent

spin-echo imaging plane (nine patients), (three

echo

sequences first echo

sagittal

time

axial plane plane

(one

of 500-800

of 20 msec.

The

709

b.

a.

Figure

1. Case

C.

Intramural leiomyoma of uterus. (a) Axial sonogram through the pelvis shows an echogenic solid mass jacent to the lower uterine segment (small arrows); this appearance does not permit determination of whether the tumor is ovarian origin. (b) Coronal Ti-weighted MR image through the pelvis shows a mass (large arrows) clearly arising from the (small arrows). (c) Sagittal T2-weighted MR image through the pelvis shows low signal intensity within this mass (arrows), uterine fibroid.

Pelvic Case

15.

in Pregnancy:

Masses

Surgical

1

Mature

of 2

right

S ummary

cystic teratoma ovary

3rd

Simple Septated

cystadenoma ovary

2nd

Mucinous

cystadenoma

2nd

9

ovary Endometnioma

unit

ovarian

mass,

sd-

irreg-

Other

facility

Peninatal

unit

cyst

Radiology

cystic

ovarian

Peninatal

unit

cystic

ovarian

Peninatal

unit

ovarian

Perinatal

unit

ovarian

department

mass Septated

of left ovary Fibrothecoma

Peninatal

mass

Corpus right Corpus

luteum ovary luteum

Corpus

luteum ovary

right

Hemorrhagic cystic ovarian mass Simple ovarian cyst

5

3rd

Simple

cyst

Radiology

2nd

Complex mass

cystic

ovarian

Peninatal

unit

Cystic

ovarian

cyst

of

2nd

Complex mass

cystic

ovarian

Perinatal

unit

Cystic

ovarian

unit

of right

ovarian

department

Paraovanian

2nd

Simple

Uterine

fibroid

2nd

Solid retrouterine

ovarian

cyst

15

Uterine

fibroid

2nd

Solid

Perinatal

unit

7 5 7.5

19

mass with wall thickening and protein. aceous fluid

Simple

8

mass

with septation mural nodule

14

4 15

cyst

7.5

of

Perinatal

23

Septated cystic ovarian mass Septated cystic ovarian mass Complex cystic ovarian mass Solid ovarian mass

cyst

Radiology department and perinatal unit

mass

13

cyst

ovarian

cyst of

left ovary 12

Simple

(cm)

15

and nodules Cystic ovarian mass, fatty mural nodule Complex ovarian cyst

Solid ovarian

of left

mass

Size

cyst with

2nd

2nd

uterus with

septations

Complex cystic mass Solid retrouterine

of left

ovary

11

Ovarian

3rd

of

left ovary

10

MR Findings

US Study

fatty

2nd

Mucinous of right

Cystadenofibroma

ovarian

2nd

5

right ovary cystic teratoma left ovary

8

cystic

id mural nodule Ovarian cyst with ular wail

Mature

7

of

Site

Cystic

3

6

Complex

wall of the consistent

an d MR Findings

Sonographic, US Findings

2nd

4

of

urgical,

mass

Mature cystic teratoma of right ovary Mature cystic teratoma

of

of

Trimester

Diagnosis

(large arrows) adof uterine versus

5

and cyst

16

Well-circumbscribed uterine mass, low

ovarian

6

ovary

16

Uterine

17

Urinary

average

pelvic

2nd

Complex

retention by retroverted uterus

2nd

Large simple

deposition

Other

pelvic

facility

Radiology

mass

fibroid

power

mass

mass

ovarian

department

Perinatal

unit

Perinatal

unit

cyst

rate of the ra-

form

for procedures.

The patient

was in-

formed that MR imaging, as a new technology, could pose unknown risks to the fetus but that no harmful effects have been found at the power levels used in the study. MR images were analyzed prospective-

with

ly, prior

710

#{149} Radiology

hospital

consent

to surgery,

with

the knowledge

8

uterine mass, low signal intensity Well-circumscribed uterine mass with low signal intensity and central cystic component Distended

retroverted

dio-frequency excitation was maintained at less than 0.4 W/kg, which represented the Food and Drug Administration’s safety guidelines for clinical MR imaging at the time this study was initiated (18). Written informed consent was obtained

use of the standard

signal intensity Well-circumscribed

that sonography

9

bladder,

0

(no mass)

uterus

had imaged

a pelvic

mass but without details of the sonographic findings. Pelvic mass lesions were assessed for size and organ of origin. Each ovarian lesion was characterized as either (a) a simple cyst if round or oval with the homogeneous signal charactenistics of water (4,12), (b) a cystadeno-

September

1990

a.

b.

Figure 2. Case 14. Subserosal leiomyoma of uterus. (a) Axial sonogram through the pelvis shows a 4 X 8-cm hypoechoic solid mass (large arrows) posterior to the uterus (small arrows); this appearance does not permit determination of whether the tumor is of uterine versus ovarian origin. (b) Axial T2-weighted MR image through the pelvis shows a well-circumscnibed solid mass of low signal intensity (large arrows). In contrast to that in Figure 3, this

mass

and

relationship

arises

from

the

posterior

to uterus

wall

permit

of the

a confident

uterus

(small

diagnosis

arrows).

Signal

of uterine

fibroid

characteristics (19).

for lesion characterization. Sonograms of four patients with pelvic masses showed that the masses were solid, but it was not possible to distinguish confidently between fibroid and solid ovarian masses. In three cases, MR images showed that these masses arose from the uterus and had low signal intensity on both Ti- and T2-weighted images (Figs 1, 2), consistent with uterine fibroids (two subserosal, one intramural). In the fourth case, the MR image showed that the mass was separate from the uterus and that it had heterogeneous high signal intensity on T2-weighted images, consistent with a solid ovarian neoplasm (Fig 3) (determined at surgery to be an ovarian fibrothecoma).

Sonograms of two patients with mature cystic tenatoma showed only complex cystic ovarian lesions suggesting cystadenoma; MR images showed fat within these masses, permitting an accurate diagnosis of matune

cystic

tenatoma

(Fig

4). The

sono-

gram of a patient with an endometrioma showed an ovarian mass suspected to be solid due to homogeneous high echogenicity; MR images showed a homogeneous mass with the signal features of diffuse hemorrhage (Fig 5), suggesting either an endometrioma on a hemorrhagic cyst. The sonognam of one patient showed a large cystic mass that pensisted

after

voluntary

voiding

and

ten urinary catheterization stetnic resident. MR helped the a. 3. Case

8.

Fibrothecoma

of left

ovary.

(a) Axial

sonogram

through

the

pelvis

shows

an echogenic mass (large arrows) that is closely related to the wall of the lower uterine segment (small arrows). (b) Axial T2-weighted MR image through the pelvis shows a well-circumscnibed mass with heterogeneous signal intensity (large straight arrows) that is clearly separate from the uterus (small straight arrows), allowing characterization as a solid left ovarian

mass.

A small

amount

of free

fluid

was

of a distended

noted

in the

pelvis

(curved

arrows).

urethra.

Limited

showed

resolution

mass”

after

tempt

at urinary

retrospect,

a successful the

by the

probably led to increased to urinary catheterization,

retroverted

other

lesions

me

rateby dalities.

as simple However,

(8).

according

to established

Additional

findings

the presence toneal fluid

criteria

tabulated

were

or absence of free intrapeniand lymphadenopathy.

Histologic proof was obtained in 16 of the 17 patients: 14 underwent elective surgery during pregnancy, one had surgi-

cal confirmation uterine

at emergency

rupture

late

one had surgical by cesarean

Volume

section.

176

surgery

in pregnancy,

confirmation For the

#{149} Number

3

and

at delivery remaining

for

Accuracy

fibroids,

in identifying

one

solid

the

ovarian

the

the catheter bladder.

RESULTS organ

neo-

plasm, two mature cystic teratomas, one endometnioma, and one distended urinary bladder. In no cases was sonography superior to MR imaging

In

uterus sistance

causing

For nine which MR hemorrhage,

imaged both rectly

at-

distortion

caused

imaging were equally sion characterization.

myosis

or adeno-

de-

second

anatomic

of origin was 100% (17 of 17) for MR imaging versus 71% (12 of 17) for sonognaphy (Table). MR imaging was superior to sonography for lesion characterization in 47% (eight of 17) of cases, which included three uter-

fibroid

imaging provided 17, Table).

images

“pelvic

catheterization.

categorized

MR (case

MR

of the

patient, finitive

serial proof

retrovertof the

repeated

ma or cystadenocarcinoma if complex without fat or diffuse hemorrhage (5), (c) a mature cystic teratoma if fat containing (6), (d) an endometnioma or hemorrhagic cyst if diffusely hemorrhagic without solid components (7), on (e) a solid ovarian neoplasm. Uterine masses were

as either

urinary

bladder due to a “trapped” ed uterus causing obstruction

b.

Figure

diagnosis

af-

by the obconfirm

resident had

gravid re-

to suspect entered

an

that empty

ovarian cystic lesions in images did not show fat on sonography and MR

as complex

accurate for Three lesions

cystic

modalities were as cystadenoma

were cysts

masses

le-

with

diagnosed con(Fig 6). Two

diagnosed with fat was

accuboth monot detect-

ed with either modality in two of four cystic teratomas, and two corpus luteum

cysts

contained

a septation,

Radiology

a #{149} 711

mural nodule, or focal wall thickening that led to a misdiagnosis of cystadenoma. The results of MR imaging changed clinical management in three cases. In one of these, a confident diagnosis based on MR findings of uterine fibroid in the second tnmesten

allowed

the

patient

to pro-

ceed to term, with myomectomy penformed at cesanean section. In the second case, confident diagnosis based on MR findings of a fibroid in the early second trimester led to cancellation of surgery; histologic diagnosis was made 8 weeks later during emergency surgery for spontaneous uterine rupture. The site of uterine rupture was not at the location of the fibroid. In the third case, a distended urinary

bladder

trapped

by

a. Figure

b.

4. Case 1. Mature cystic teratoma of right ovary. (a) Axial Ti-weighted MR image at level of kidneys shows a large septated cystic mass with a central component of high signal intensity (arrow), consistent with fat. (b) Axial T2-weighted image at same level as a. The central component (arrow) remains isointense to fat, while the major contents of the cyst re-

main

isointense

to urine.

Internal

chemical

shift

artifact

was better

seen

on adjacent

images.

a retro-

verted uterus had been misinterpreted as a cystic mass at sonography, as discussed above; a diagnosis based on MR findings permitted cancellation of scheduled surgery. A small amount of free fluid was noted on MR images in only four patients,

one

each

with

cystadenoma,

fibrothecoma, leiomyoma, and conpus luteum cyst. With MR imaging, lymphadenopathy was correctly excluded in all 16 cases with surgical confirmation. DISCUSSION The pregnant patient with a pelvic mass presents a unique diagnostic problem that may not be adequately solved with sonography (4,12). Precise characterization of the pelvic mass is essential to either plan surgery in the pregnant patient or confidently post-

pone

surgery.

In our series,

sonographic

find-

ings in 44% (seven of 16) of pregnant patients with pelvic masses. The current study adds several important observations to the report by

Weinreb

et a!. First,

b.

Figure

5. Case

the pelvis

9.

Endometrioma

shows

T2-weighted

a mass

image

signal

intensities

cystic

lesion,

of left

of high

shows

marked

on Ti- and suggesting

signal

ovary.

intensity

hypointensity

T2-weighted

images

endometrioma

we found

that

MR

imaging can supplement information obtained with both endovaginal and transabdominal sonography. Second,

simple corpus luteum cyst at sonography were not studied with MR imaging. Corpus luteum cysts have been descnibed on MR images as round or oval masses with imperceptible or thin walls, having homogeneous low signal intensity on Ti-weighted images and high signal intensity on T2-weighted images, consistent with “simple” fluid (12). In our series, two of three corpus luteum cysts appeared complex on both

MR and

sonographic

images,

raising

the suspicion of a cystadenoma. study may possess selection

bias

Our in fa-

von of atypical

population. Whereas the lesions in our series were plasms (eight of 17 cases),

peared to be simple cysts at sonography were less likely to be referred for MR

in the prior

since

most common ovarian neothe majority

study

were

.

Radiology

corpus

patients

with

luteum masses

cysts, that

ap-

imaging.

Mature

con-

pus !uteum cysts and fibroids (12). This may partially reflect different entry cnitenia into our series, since masses smallen than 6 cm that met the criteria for a

712

Ti-weighted

(arrow)

posterior

of the

mass

MR

to gravid

(arrow).

characterizes

or hemorrhagic

we have demonstrated the usefulness of MR imaging in a different patient

of lesions

(a) Sagittal

This

the mass

image

uterus.

through

(b) Axial

combination

of

as a hemorrhagic

cyst.

MR imag-

ing supplemented information obtamed with sonography in 47% (eight of 17) of the patients. Similarly, a recent study by Weinreb et a! (12) performed at 0.35 T found that MR imaging contributed additional information

to transabdominal

a.

teristic

of these sistent shift

cystic features

tenatomas on

features, with

artifacts,

MR

signal

fat and were

have images

charac(6). Two

intensity

internal demonstrated

con-

chemical

two

cases, (Fig

of mature

cystic

allowing

confident

4). In another noted signal with fat in their

teratoma

series, the intensity single case

in pregnancy

(12); failure to demonstrate internal chemical shift may have been related to the lower field strength of their magnet (0.35 T), since the chemical shift antifact is more prominent at higher field strengths if gradient magnitudes are not increased proportionately (20). In our series, failure to detect fat with MR imaging in two of these tumors corre-

lated with evaluation.

a paucity of fat at histologic These two cases may repre-

sent uncommon presentations, since recent report noted signal intensity consistent with fat in 22 of 23 cystic

atomas (6). Ovarian cystadenomas variety thus

(5,10). in

of four

diagnosis investigators consistent

sible

may

of appearances limiting

their

Furthermore, to distinguish

on MR

have

a ten-

a

images,

characterization

it may them

not from

September

be poscystade-

1990

In conclusion,

sonography

should

ne-

main the primary imaging tool in pregnant women who present with pelvic masses. If results of sonography are equivocal, MR imaging can provide supplemental information that may influence patient treatment. U

References 1.

Buttery BW, Beischer NA, fee CAJ. Ovarian tumors J Aust 1973; 1:345-349.

2.

White

3.

Am J Obstet Gynecol Novak ER, Lambrou ian

KC.

tumors

1975;

4. a.

b.

Figure

6. Case

through suggesting

6.

Mucinous

the pelvis simple

ial T2-weighted

cystadenoma

of left

ovary.

Axial

(a)

Ti-weighted

MR

image

shows a homogeneous left pelvic mass (arrow) with low signal intensity, fluid. The top of the uterus is anterior and to the right of the mass. (b) Aximage

tensity consistent with this lesion as a complex

at same

level

as a. Although

simple cystic

fluid mass.

contents,

the

mass

(arrow)

the presence

displays

of multiple

high

septations

signal

5.

in

Lavery Gumpel

6.

tentiab

ability

of MR

imaging

to depict

lymphadenopathy may help in the detection of malignancy. However, since none of our patients demonstrated enbarged lymph nodes on MR images or at surgery, the sensitivity of MR imaging for lymph node detection could not be assessed in this study. In women presenting with a pelvic mass early in pregnancy, MR imaging should be postponed until the second

trimester most

for several

common

reasons.

lesion

First,

occurring

the

in the

first trimester is a corpus luteum cyst. Since most of these can be shown with sonography to regress by the end of the first

trimester

(4),

the

expense

of MR

imaging will be avoided in many cases. Second, operative intervention is usuably delayed until the second trimester, since

the

rate

of postoperative

death falls from 35% in the first ten to 2% in the second trimester this

difference

reflects

both

fetal trimes-

(1);

anesthesia-

related fetal death and spontaneous abortion (most common in the first tnmester). Third, although there is no evidence to suggest that MR imaging is hazardous to the embryo at the magnetic field strength and radio frequency used

for

clinical

MR

imaging,

the

first trimester represents the major penod of organogenesis. Thus, the British National Radiological Protection Board has suggested that “it might be prudent to exclude pregnant women during the first trimester” (21). Scientists at the National Institutes of Health Consensus

Development

imaging as with

Volume

Conference

on

MR

agreed, suggesting that “MRI, all interventions in pregnancy,

176

#{149} Number

3

should be used during the first tnimester only when there are clean medical indications and it offers a definitive ad-

vantage over other tests” (22). MR imaging of the pregnant

patient

In nonpregnant MR imaging

I I. 12.

patients, the role of in pelvic mass lesions in-

13.

masses:

Togashi 1987; Zawin

(6), but surgery

cy remains tial obstetric

during

indicated due complications

ovarian mass. Demonstration rhage within a cystic lesion image allows characterization

evidence

of bleeding

a homogeneous

ance

without

solid

endometnioma.

on a suspected

during suggested

pregnancy, both

components

prior

serial to confirm

to labor

and to exclude interval unsuspected hemorrhagic

(5),

16.

19.

and

MS. Hricak

at

differentiation

between

adeno-

H.

Magnetic

resonance

anomalies in utero: early cxAJR 1985; 145:677-682. SM, Stark DD, Filly RA, Callen PW. H, Higgins CB. Obstetrical magnetic

imaging:

maternal

anatomy.

Radi-

154:421-425. SM, Filly RA, Stark DD, et a!. Obmagnetic resonance imaging: fetal anatom’. Radiology 1985; 154:427-432. Weinreb JC, Lowe 1W, Santos-Ramos R, Cunningham FG, Parkey R. Magnetic resonance imaging in obstetric diagnosis. Radiology 1985; 154:157-161. Cohen JM, Weinreb JC, Lowe TW, Brown C. MR imaging of a viable full-term abdominal pregnancy. AJR 1985; 145:407-408. Kido OK, Morris TW, Erickson JL, Plewes DB, Simon JH. Physiologic changes during high field MR imaging. AJNR 1987; 8:263-266. Weinreb JC, Barkoff ND, Megibow A, Demo-

1985;

R. AiR

Soila

KP, shift

resonance 820. National

cal

is of

F.

detection

McCarthy stetrical

ical

22.

L, Comite and

of fetal

McCarthy

sure

endometnioma sonography regression

Golbus

nate.

21.

Ovani-

Radiology

leiomyoma with MR imaging. Radiologv 1989; 171:531-534. Arrive L, Hricak H, Martin MC. Pelvic endometriosis: MR imaging. Radiology 1989; 171:687-692. Mawhinney RR, Powell MC, Worthington BS, Symonds EM. Magnetic resonance imaging of benign ovarian masses. Br J Radiol 1988; 61:179-186. McCarthy S. MR imaging of the uterus. Radiology 1989; 171:321-322. Weinreb JC, Brown CE, Lowe TW, Cohen JM, Erdman WA. Pelvic masses in pregnant patients: MR and US imaging. Radiology 1986; 159:717-724. McCarthy SM, Filly RA, Stark DO, Callen PW.

The

tinguishing vic masses 20.

K, et al.

S. Scoutt

appearance

and

vised

pen-

Radiology

MR imaging.

uterus:

poulos

favors

is not

CT correlation.

teratomas:

resonance

appear-

If surgery

formed

the mass

on MR images

hemorrhagic

Gyne-

et al. observations at CE,

K, Itoh

mvosis

ology

15.

18.

sion as probably an endometnioma on hemorrhagic cyst (7), which might not require surgery due to the possibility of spontaneous regression later in pregnancy. Although both benign and malignant ovarian neoplasms may exhibit

Surg

K, Nishimura

larged

Hricak

17.

of hemonon an MR of a le-

Ovar-

Gynecol

MR imaging. Radiolog’ 1989; 171:693-696. Togashi K, Ozasa H, Konishi I, et al. En-

perience.

14.

pregnan-

to the potenof an

JO.

Obstet

MR imaging

US and

162:669-673. M, McCarthy

imaging

determinate at sonognaphy has been demonstrated (19). In cases in which sonograms show that a lesion is clearly ovarian in origin, the role of MR imaging is more limited. Demonstration of fat on MR images allows chanactenization of the lesion as a mature cystic ter-

atoma

9.

10.

is particularly valuable when sonograms are unable to show whether a pelvic mass represents a fibroid or an ovarian lesion, since a confident diagnosis of a uterine fibroid may eliminate the need for surgery during pregnancy.

pregnancy.

1987; 162:319-324.

8.

po-

in

adnexa during early pregnancy. col Obstet 1986; 163:319-323. Mitchell DG, Mintz MC, Spritzer

an cystic

(5). The

MacaMed

116:544-550.

CD. Woodruff

pregnancy.

Endometriosis:

on MR images

tumors

1973;

DW,

pregnancy.

JP, Koontz WL, Layman L, Shaw L, U. Sonographic evaluation of the

1.5 1, with

7.

nocarcinoma

in

46:401-406.

Adnexal

in-

categorizes

Ovarian

Fortune

value

of MR imaging

leiomyomas from when sonographv

1990; Viamonte

M, Starewicz

misregistration

effect

Radiology

Radiological guidance

imaging.

pcI-

154:295-299.

imaging.

during

in dis-

other solid is indetermi-

on

PM. in

Protection acceptable

Chem-

magnetic

1984;

153:8

Board. limits

of

nuclear magnetic resonance Br J Radiol 1983; 56:974-977.

Abrams HL, Berne AS, netic resonance imaging.

Dodd GD, National

19Re-

expodin-

et al. MagInstitutes

of Health Consensus Development Conference Statement. Vol 6, no. 14. Bethesda, Md: National

Institutes

of Health,

1987.

delivery

growth in an neoplasm.

Radiology

#{149} 713