Clinical Radiology (1991)43, 197 204

The Role of Magnetic Resonance Imaging in the Assessment of

Local Recurrent Breast Carcinoma
H. G. LEWIS-JONES, G. H. W H I T E H O U S E and S. J. L E I N S T E R
Liverpool University Magnetic Resonance Research Centre

Experience with magnetic resonance imaging (MRI) of the breast remains limited. MRI studies
to date have shown that differentiation of carcinoma from certain benign breast changes can be
difficult.
The problem of suspected tumour recurrence in patients with known but treated breast
carcinoma is considered. Forty-five patients were studied, all having been treated by
lumpectomy combined with radiotherapy and/or chemotherapy. Suspicion of recurrence was
suggested by X-ray mammography or clinically by the presence of a recurrent breast mass,
breast pain, or nipple discharge. The principle differential diagnosis rested between post-
treatment fibrosis and recurrent tumour. Axial and sagittal images were obtained using T1- and
T2-weighted pulse sequence. Images were enhanced with intravenous gadolinium D T P A in
cases where there was a mass.
The tomographic format and inherent high soft tissue contrast provided by M R I are of
particular value in this situation. The morphological appearances of recurrent tumour, fibrosis,
and other p0st-radiation affects are described and compared. MRI allowed accurate
differentiation in the majority of case. In equivocal cases enhancement of mass lesions with
gadolinium D T P A provided excellent confirmatory evidence of recurrent turnout. Lewis-Jones,
H.G., Whitehouse, G.H. & Leinster, S.J. (1991). Clinical Radiology 43, 197-204. The Role of
Magnetic Resonance Imaging in the Assessment of Local Recurrent Breast Carcinoma

Early studies of breast MRI have shown considerable
overlap in the appearances of breast carcinoma, breast
dysplasia and benign breast disease (Ross et al., 1982; E1
Yousef et al., 1984; Heywang et al., 1987). X-ray
mammography has been shown to be more sensitive than
MRI in the detection of carcinoma (Turner et al., 1988)
and for these reasons no clinical application for breast
MRI has been demonstrated so far. The advent of
lumpectomy and radiotherapy with breast conservation
in the treatment of breast cancer (Montague et al., 1979,
Fisher et al., 1985) has brought with it the problem of
suspected local tumour recurrence where clinical examin-
ation and conventional imaging are often equivocal due
mainly to post-treatment fibrosis.
PATIENTS AND M E T H O D S
Forty-five female patients (age range 27-81 years) with
suspected recurrent breast carcinoma were examined
prospectively by MRI. Forty-two had been treated by
lumpectomy combined with radiotherapy and three by
lumpectomy alone. Four patients had received or were
currently being treated by adjuvant hormonal or chemo-
therapy. Recurrence was suspected either following rou-
tine biannual X-ray mammography (nine patients) or
clinically by the presence of a palpable mass (27 patients),
local breast pain (seven patients) or a nipple discharge
(two patients). Mammography had been performed
within the four weeks prior to MRI in 34 of the cases and
in the remaining 11 cases referral was direct to MRI,
rnammography having been performed routinely on a 6
monthly basis but not immediately prior to M R I itself.
Correspondence to: Dr H. G. Lewis-Jones, Magnetic Resonance
ResearchCentre, LiverpoolUniversity,PembrokePlace,P.O. Box t47,
LiverpoolL69 3BX.
A General Electric Signa 1.5 Tesla system was used for
the MRI investigations. For breast examinations the
patient lay prone in the magnet bore with the breast
suspended in a purpose built holder containing a 5 inch
general purpose surface coil. A slice thickness of 5 mm
with a 2 mm gap was used. Tl-weighted axial sections
using a spin echo pulse sequence (TR-500 ms, TE-20 ms)
and T2-weighted and proton density sagittal sections also
using a spin echo pulse sequence (TR-2000 ms, TE-20 and
80 ms) were performed initially. These images were
carefully inspected and, if a mass was demonstrated in the
breast, enhancement with gadolinium D T P A was per-
formed. Sequential fast scan single slice images at the
relevant location were repeated at the rate of approxima-
tely two per minute for 8 rain following injection of
gadolinium. The single slice images obtained were gra-
dient echo images (GRASS) using a T R of 200 ms, a TE of
13 ms and a flip angle of 90 ° to provide a 'T1 type' fast
scan. Finally the initial Tl-weighted axial sequence was
repeated, the entire MR study lasting approximately 40
min.
The MR images were reported with only the clinical
information and history available, except where referral
had been due to abnormal findings at routine mammogra-
phy in which case the relevant mammogram was
inspected following the initial interpretation of the spin
echo images.
Interpretation was based on our own initial experience
as well as other information available in the literature (El
Yousef et al., 1984, 1985; Dash et al., 1986; Heywang et
al., 1986; W e i n e r et al., 1986; Stelling et al., 1987).
Histological confirmation of diagnosis was obtained in
those patients where clinical suspicion of recurrent malig-
nancy was high or where imaging itself suggested recur-
rence. Histological diagnosis was obtained in 24 of the 45
cases studied. In the remaining 21 cases close clinical
198 CLINICALRADIOLOGY

follow up at 3 monthly intervals with shared care l~etween

surgeon and radiotherapist was instituted and stable
follow up defined as no clinical change at four successive 3
monthly follow up visits, i.e. at 12 months judged by serial
clinical examination of the breast (minimum follow up 12
months, maximum follow up 24 months, mean follow up
15.6 months).

RESULTS

The overall study results are shown in Table 1. In 13

cases M R I was interpreted as showing recurrent tumour.
There were 11 cases of proven tumour recurrence and all
were correctly identified by M R I leaving two false
positive studies. Recurrent tumour showed predictable (a)
morphological appearances with masses of varying size
(minimum 1.2 cm, maximum 4.6 cm). In all cases of
recurrence the masses were heterogenous with interme-
diate signal close to that of muscle on T 1-weighted images
(Figs la and 2a) and slightly higher signal than muscle on
T2-weighted sequences though not as high as adjacent fat
(Figs l b and 2a). The T2-weighted images allowed the
small tumour masses of higher signal to be distinguished
from areas of glandular or dysplastic breast tissue as these
remained at low signal levels relative to muscle. Ten of the
11 recurrent tumours demonstrated irregular spiculated
borders with distortion of breast architecture suggesting
local infiltration by tumour although this was not a
specific sign of recurrence as it was also present in four
fibrous masses, one abscess and one haematoma. Histolo-
gical confirmation was obtained in 10 cases and in one
there was clinical evidence of local tumour growth and
disseminated metastatic disease resulting in death six
months following the scan. There were two false positive (b)
studies where M R wrongly indicated recurrent tumour.
In the first of these, histology showed a small area of
lobular hyperplasia and close clinical follow up was
advised in view of a possible premalignant potential. In
the second case, where tissue characterization studies
were most suggestive of recurrent tumour, histology
revealed a mixture of fat necrosis and fibrous tissue_
Interestingly, there was no evidence of significant en-
hancement with gadolinium in this particular case. In
terms of detection of recurrent tumour there were no false
negative examinations, M R ! successfully detecting all 11
cases of recurrent breast carcinoma in this series.
Histology in nine of the cases of recurrence showed
ductal adenocarcinoma of varying degrees of differentia-
tion. The remaining case of histology proven recurrence
was due to mucinous adenocarcinoma and as would be
expected this lesion appeared smooth and lobulated on
M R I with a well defined border and minimal architec- (c)
Fig. 1 - Patient presenting with breast pain and generalized thickening
and lymphoedema of the medial half of the breast 4 years after treatment
Table 1. Correlation of MR results with histology and clinical outcome for breast cancer. (a) The Tl-weighted axial image shows a heteroge-
nous mass (arrow), fixed posteriorly to the pectoralis muscle associated
Number of patients with M R l features of: with considerable oedema of the dermis. (b) Images from the sagittal
V E M P sequence with the proton density image on the left and the T2-
Benign post- Other weighted image on the right. The T2-weighted sagittal image shows the
Recurrent Fibrous radiation benign mass (arrow) to be of higher signal than intercostal muscle which
tumour mass effects masses supports the morphological appearances of t u m o u r recurrence. (c) The
craniocaudal m a m m o g r a m reveals the spiculated mass (arrow) beneath
True positive 11 10 11 9 the skin surface which would itself indicate recurrent tumour. There was
True negative 32 35 34 36 no evidence of microcalcification.
False positive 2 0 0 0 The patient declined biopsy or further surgery but over a 4 week
False negative 0 0 0 0 period the mass clinically increased in size with severe tethering and local
oedema confirming the presence of recurrent tumour.
 MRI IN THE ASSESSMENT OF RECURRENT BREAST CARCINOMA 199

(~)

@) (c)

Fig. 2 Recurrent t u m o u r in a patient 3 years following treatment. (a) A lobulated mass (arrow) close to the chest wall is shown, with an intermediate
signal on the proton density image (left) while the T2-weighted image (right) shows uneven signal with areas of high signal relative to muscle. These
signal changes are typical of recurrent tumour. (b) Axial fast scans through the turnout before (above) and after (below) intravenous gadolinium DTPA.
There is patchy but marked enhancement in parts of the t u m o u r (arrow). (c) Lateral m a m m o g r a m showing a poorly defined area of increased density
posteriorly in the breast (arrows) which was asymmetrical when compared with the contralateral breast. There is minimal architectural distortion with
no evidence of microcalcification.
A breast biopsy revealed recurrent poorly differentiated ductal adenocarcinoma.

tural distortion (Fig_ 3a). The tissue characterization
studies, however, were in keeping with tumour (Fig. 3b)
rather than benign disease and given knowledge of
previous histology no error should be made.
In 10 cases masses were shown whose morphology
suggested a recurrent tumour mass while tissue character-
ization studies showed a signal intensity lower than
muscle on T 1- and T2-weighted images indicating that t h e
masses were more likely due to fibrosis (Figs 4a, b).
Histological confirmatf0n was obtained in three cases
while in seven close clinical observation over a mean
period of 17 months remained stable.
Other post-treatment effects were demonstrated in 11
cases. No underlying mass was shown in these cases but
considerable distortion of breast architecture related to
lumpectomy scars, tenting of the underlying pectoralis
muscle (Fig. 5) and post-radiation skin thickening with
localized breast oedema were observed. Clinical follow up
remained stable in all 11 cases over a mean period of 15
months.
Nine other benign masses were shown including six
haematomas (Fig. 6), one abscess and two simple cysts
(Fig. 7). Histological confirmation by aspiration or
surgery was obtained in all but two small haematomas
which remained stable at clinical follow up with no
increase in size.
200 CLINICAL RADIOLOGY

N i n e cases were referred following r o u t i n e follow up X-

r a y m a m m o g r a m s which were r e p o r t e d as suspicious o f
recurrence. There were o n l y two cases o f p r o v e n recur-
rence in this g r o u p a n d m a m m o g r a p h y had shown a
distinct mass c o n t a i n i n g microcalcification in only one. In
the r e m a i n i n g seven cases M R revealed benign post-
t r e a t m e n t effects which a c c o u n t e d for e r r o n e o u s interpre-
t a t i o n o f m a m m o g r a m s . F i g u r e 5(b) illustrates an a p p a r -
ent m a s s on m a m m o g r a p h y while M R d e m o n s t r a t e d only
severe scarring a n d could confidently exclude an underly-
ing mass. M a m m o g r a p h y was u n a b l e to distinguish
between masses due either to to r e c u r r e n t t u m o u r or
residual p o s t - t r e a t m e n t fibrosis. Benign masses, p a r t i c u -
larly the two cases o f simple cysts were correctly identified
by m a m m o g r a p h y b u t p o s t - o p e r a t i v e h a e m a t o m a s gave
rise to masses which c o u l d n o t confidently be labelled
benign in f o u r out o f six cases. I n a further four cases
m a m m o g r a m s revealed masses in which the t o m o g r a p h i c
f o r m a t o f M R I revealed severe scarring b u t with no true (c)
mass. O f the 34 cases where m a m m o g r a p h y h a d been Fig. 3 - Recurrent t u m o u r 1 year following treatment. (a) Well-defined
p e r f o r m e d within f o u r weeks o f M R I there were 21 cases Iobulated mass (arrow) demonstrated on Tl-weighted axial images. (b)
where m a m m o g r a p h y was inconclusive. The typical tissue characterization findings of recurrent tumour with
intermediate signal on the proton density image (left) and high signal
relative to muscle on the T2-weighted image (right). The original tumour
histology revealed a mucinous adenocarcinoma which explains the
apparent benign appearances of recurrent turnout in this case. (c) The
craniocaudal mammogram illustrates the well defined lobular mass in
the atrophic breast. The margins are crisp in places while a little blurred
in other areas (small arrows). Given the original histology and previous
mammograms, recurrent tumour would be most likely_ This was
confirmed histologically.

E n h a n c e m e n t with g a d o l i n i u m D T P A in the situation

o f suspected recurrence p r o v i d e d excellent supportive
evidence o f recurrent t u m o u r , m a r k e d e n h a n c e m e n t
being d e m o n s t r a t e d in all 10 recurrent t u m o u r s . False
positive e n h a n c e m e n t occurred in only one case where
s u b s e q u e n t h i s t o l o g y revealed an area o f a t y p i c a l l o b u l a r
h y p e r p l a s i a . In seven o f 11 r e c u r r e n t turnouts enhance-
m e n t was n o n - h o m o g e n o u s , some lesions showing non-
e n h a n c i n g areas centrally a n d p e r i p h e r a l l y which corres-
p o n d e d to fibrotic areas associated with r e c u r r e n t t u m o u r
(Fig. 8). This in itself m a y be o f value to the surgeon
seeking histological p r o o f o f recurrence as the relatively
(a) v a s c u l a r e n h a n c i n g areas w o u l d be m o r e likely to p r o v i d e
n e o p l a s t i c tissue at biopsy. Seven fibrous masses and
three h a e m a t o m a s showed no evidence o f significant
e n h a n c e m e n t with g a d o l i n i u m c o n t r a s t i n g well with the
r e c u r r e n t t u m o u r (Fig. 9).

DISCUSSION

N o r m a l b r e a s t tissue has an extremely v a r i a b l e c o m p o -

sition related to the a m o u n t s o f b r e a s t s t r o m a , duct
systems a n d the h o r m o n a l l y d e p e n d e n t p e r i d u c t a l tissues.
T h e presence a n d variety o f benign b r e a s t disease, as in
m a m m o g r a p h y , m a k e s it difficult to distinguish t u m o u r s
f r o m dense dysplastic tissue a l t h o u g h o c c a s i o n a l l y T2
d o m i n a n t images can reveal a t u m o u r with higher signal
t h a n s u r r o u n d i n g dysplasia. H o w e v e r , certain benign
b r e a s t diseases can display a higher signal on T2-weighted
images as can some f i b r o a d e n o m a s . T h e i n a b i l i t y o f M R
to d e m o n s t r a t e microcalcification places it at a further
d i s a d v a n t a g e when c o m p a r e d with m a m m o g r a p h y and,
(b) given these facts, it is difficult to envisage a p r i m a r y role
 MRI IN THE ASSESSMENT OF RECURRENT BREAST CARCINOMA 201

(a)

(b) (c)
Fig. 4 Fibrotic mass demonstrated posteriorly in the breast 2 years after treatment. (a) Tl-weighted axial image shows low signal mass with spiculated
borders posteriorly in the breast (arrow) adjacent to pectoralis muscle. (b) The T2-weighted sagittal image shows the mass (arrow) to be of low signal on
this sequence also suggesting fibrosis rather than recurrent tumour_ (c) The lateral m a m m o g r a m shows a mass close to the chest wall with a spleulated
border, architectural distortion and the appearances of recurrent tumour.
A subsequent breast biopsy revealed no neoplastic tissue and the patient has remained stable at 24 m o n t h follow up.

for breast M R I at this current stage. In assessing patients
for recurrent tumour there were a number of advantages
which could offset some of the pitfalls of M R I in the
breast. Firstly the site of operation was known so the
examination could be targeted to the relevant area. In
addition we had prior knowledge of the patients previous
histology and this proved to be of particular value in cases
ofrecurrent mucinous adenocarcinoma where morpholo-
gically the recurrent tumour would appear benign with
well-defined lobulated borders.
The study demonstrated good correlation between the
results of M R I and subsequent histology or close clinical
follow up. M R I successfully detected all 11 recurrent
tumours giving 100% sensitivity but two false positive
studies resulted in a lower specificity of 94% although it
should be noted that the number of patients in this study
is too small for these calculations to be valid.
Recurrent tumour most commonly appeared as an
irregular mass with radiating spicules associated with
distortion of breast architecture. Secondary signs of
carcinoma such as skin thickening, skin retraction and
nipple invertion were well demonstrated but of limited
202 CLINICAL RADIOLOGY

Fig. 5 - (a) Tl-weighted axial image shows outdrawing and tenting of

the pectoralis muscle associated with low signal areas of fibrosis. There
does not appear to be a true mass as had been suggested by the
mammogram. (b) The lateral mammogram reveals an apparent spicu-
lated mass (arrows) close to the chest wall with deformity and scarring of
the skin surface at the site of the previous surgery. The three discrete
areas of calcification were not related to the mass on the craniocaudal
view and were secretory in nature. There appears to be distortion of
anatomy and although no microealcification could be demonstrated the
appearances wer e regarded as due to recurrence.
The appearances remained unaltered over an 18-month follow-up
period and there was no change on clinical examination.
(a)

(b)
Fig. 6 - (a) Axial MR image showing large haematoma with 'mother of
pearl' appearance surrounded by low signal fibrous capsule. Tissue
characterization studies revealing high signal on T1- and T2-weighted
images confirmed that the mass was a haematoma. (b) The correspond-
ing m a m m o g r a m also demonstrates a dense spherical smooth walled
mass although not all its borders are completely distinct and crisp. The
appearances were felt on balance to be due to a large haematoma but
recurrence could not be excluded without biopsy.
(a) The mass was aspirated and chronic haematoma aspirated.
 MRI IN THE ASSESSMENT OF RECURRENT BREAST CARCINOMA
Fig. 7 - T 2 - w e i g h t e d sagittal image shows smooth lobulated mass
beneath lumpectomy scar with very high homogenous signal. The low
signal demonstrated on the T 1-weighted image confirmed that this was a
simple cyst.
Fig. 8 - Tl-weighted axial image showing rim enhancement of a
recurrent t u m o u r mass (arrow) before (above) and after (below)
intravenous gadolinium DTPA. The fibrous centre of the t n m o u r fails to
enhance. Same case as illustrated in Fig 1 (a) and (b).
value as these changes could also be produced by
radiation or post-surgical scar tissue. T u m o u r was best
differentiated from scar tissue on the T2-weighted mul-
tiple echo sequences where the tumour was of higher
signal in relation to dysplastic tissue and fibrosis.
203
Fig. 9 Fast Grass scan before and after gadolinium D T P A demon-
strating a low intensity fibrotic mass just beneath the dermis (arrow)
which does not enhance with gadolinium Note the enhancing glandular
breast tissue in other areas of the breast which is a normal finding in this
27-year-old patient.
Fibrotic masses, in m a n y respects appeared morpholo-
gically similar to recurrent tumour and again it was the
low demonstrated on T2-weighted images that was most
reliable in distinction from tumour recurrence. Studies of
similar histology in other sites such as the pelvis and
mediastinum have shown that fibrous masses have longer
T1 values relative to muscle and shorter T2 values (Ebner
et al., 1988; N y m a n et al., 1989), the same pattern of
disease being shown in this study.
Benign post-treatment effects were clearly demon-
strated at M R I and although m a n y of these M R I
appearances corresponded well with equivocal abnor-
malities discovered at X-ray m a m m o g r a p h y which could
be misinterpreted as indicating malignancy, the tomo-
graphic format and high soft tissue resolution of M R I
allowed confident exclusion of any underlying mass.
Scarring following surgery and post-radiation effects
created difficulty in the interpretation of X-ray m a m m o -
grams, the majority of reports being equivocal. Overall
there was a poor correlation of m a m m o g r a p h y with
subsequent histology and clinical outcome although the
presence of microcalcification in the region of the lumpec-
tomy scar as identified in three of the cases of recurrence
provides irrefutable evidence of recurrent tumour and of
course is not visible at MRI.
No examples of diffusely infiltrating recurrent tumour
were present in this study and in this situation M R I has a
potential weakness. The diffuse low signal intensity
change seen in such cases on Tl-weighted images would
render them indistinguishable from diffuse m a m m a r y
dysplasia.
Gadolinium enhancement in this clinical situation
provided excellent confirmatory evidence of recurrent
tumour and allowed more confident differentiation of
tumours from fibrous masses and other post-treatment
effects. There was only one case where enhancement was
not associated with tumour and histology here revealed
an area of atypical lobular hyperplasia, possibly prema-
lignant in its own right.
The clinically evaluated probability of recurrence,
assessed as high, moderate or low, was correlated with the
204 CLINICAL RADIOLOGY
p r o v e n o u t c o m e a n d m o d e o f referral. This revealed a
wide spread o f cases across the range of clinical risk. Only
five o f 13 cases of recurrence were evaluated as high risk
a n d in a further seven cases high risk evaluation was
associated with a benign outcome, suggesting that clinical
evaluation is indeed difficult in the p o s t - t r e a t m e n t breast
of p o o r predictive value c o m p a r e d to M R I .
It is believed that M R I with its t o m o g r a p h i c f o r m a t
a n d high soft tissue detail can clearly d e m o n s t r a t e masses
within the p o s t - t r e a t m e n t breast a l t h o u g h differentiation
of recurrent t u m o u r from p o s t - t r e a t m e n t fibrous masses
can be difficult on m o r p h o l o g i c a l g r o u n d s alone. Tissue
characterization studies proved helpful in differentiating
these two conditions while in equivocal cases enhance-
m e n t with g a d o l i n i u m D T P A in this particular clinical
situation provided excellent c o n f i r m a t o r y evidence of
recurrent turnout.
Currently M R I is costly a n d n o t widely available.
M a m m o g r a p h y a n d u l t r a s o u n d c o m b i n e d with fine
needle aspiration when indicated are likely to r e m a i n the
initial investigations of choice. W e believe that gadoli-
n i u m e n h a n c e d M R I of the breast should be employed in
those cases where initial m a m m o g r a p h y has d e m o n -
strated a mass a n d s u b s e q u e n t fine needle aspiration or
biopsy has been negative for t u m o u r . The biopsy itself
m a y cause some a b n o r m a l appearances b u t M R I still has
the ability to detect recurrent disease a n d in this way m a y
have a distinct role to play in the assessment of patients
with suspected recurrent breast carcinoma.
REFERENCES
Dash, N, Luptein, AR, Daffner, RH, Deeb, ZL, Sefczek, RJ & Schapiro,
RL (1986). Magnetic resonance imaging in the diagnosis of breast
disease. American Journal of Radiology, 146, 119-125.
Ebnor, F, Kressel, HY, Mintz, MC, Carlson, JA, Cohen, EK, Schiebler,
M, Gefter, W & Axel, L. (1988). Tumour recurrence versus fibrosis
in the female pelvis: differentiation with MR imaging at 1.5T.
Radiology, 166, 333 340.
E1 Yousef, SJ, Duchesneau, RH, Alfidi, RJ, Haaga, JR, Bryan, PJ &
LiPuma, JP (1984). Magnetic resonance imaging of the breast.
Radiology, 150, 761 766.
E1 Yousef, SJ, O'Connell, DM, Duchesneau, RH, Smith, MJ, Hubay,
CA & Guyton, SP (1985)_ Benign and malignant breast disease;
magnetic resonance and radiofrequency pulse sequences. American
Journal of Radiology, 145, 1 8.
Fisher, B, Bauer, M, Margolese, R, Poisson, R, Pilch, Y, Redmond, C et
al. (1985). Five year results ofa randomised clinical trial comparing
total mastectomy and segmental mastectomy with or without
radiation in the treatment of breast cancer. New England Journal of
Medicine, 312, 665 673.
Heywang, SH, Basserman,R, Fenzl, G, Nathrath, W, Hahn, D, Beck, R
et aL (1987). MRI of the breast-histopathologic correlation. Euro-
pean Journal of Radiology, 7, 175-182.
Montague, ED, Guttierrez, AE, Barker, JL, Tapley, ND & Fletcher,
GH (1979). Conservative surgery and irradiation for the treatment
of favourable breast cancer. Cancer, 43, 1058-1061.
Nyman, RS, Rehn, SM, Glimelius,BLG, Hayberg, HE, Hemmingsson,
AL & Sundstrom, CJ (1989). Residual mediastinal masses in
Hodgkin disease: prediction of size with MR imaging. Radiology,
170, 435-440_
Ross, RJ, Thompson, JS, Klm, K & Bailey, RA (1982). Nuclear
magnetic resonance imaging and evaluation of human breast tissue:
preliminary clinical trials. Radiology, 143, 195-205.
Stelling, B, Powell, DE & Mattingly, SS (1987). Fibroadenomas:
histopathologic and MR imaging features. Radiology, 162, 399-407.
Turner, DA, Alcorn, FS, Shorey, WD, Stelling, CB, Mategrano, VC,
Mertern, CW et al. (1988). Carcinoma of the breast: detection with
MR imaging versus xeroammography. Radiology, 168, 49-58.
Weiner, JI, Chako, AC, Merten, CW, Gross, S, Coffey, EL & Stein, HL
(1986). Breast and axillary tissue MR imaging: correlation of signal
intensities and relaxation times with pathologic findings.Radiology,
160, 299-305.