B re a s t M R I m a g i n g

Atlas of Anatomy, Physiology,

Pathophysiology, and Breast Imaging
Reporting and Data Systems Lexicon
Sarah R. Martaindale, MD

KEYWORDS
 Breast MR Imaging  BI-RADS  Anatomy  Atlas

KEY POINTS
 The latest edition of the MR imaging Breast Imaging Reporting and Data Systems lexicon brings
consistency to the descriptors used across the lexicon and eliminates underused or confusing
terminology.
 A more cohesive set of descriptors aids in lesion management and ultimately improves patient care
by facilitating communication between radiologists and referring physicians.
 A new section on breast MR imaging implant evaluation assists radiologists in identifying key signs
of implant rupture and effectively communicating those findings to surgeons.

INTRODUCTION surgical planning. Components of the breast

include the skin, superficial fascia, deep fascia,
In 1993, the first edition of the Breast Imaging breast parenchyma, and the nipple–areola com-
Reporting and Data System (BI-RADS) was plex. The breast parenchyma is composed of a
released with the goal of standardizing mammog- glandular epithelium, fibrous stroma, and fat. The
raphy reporting and interpretation.1 Ten years later glandular epithelium is composed of 15 to 20
in 2003, MR imaging was added to the fourth edi- lobes, which are in turn composed of lobules or
tion of the BI-RADS lexicon to provide consistency terminal ductules, the milk-producing glands.
in communication of findings across modalities.2 Each breast lobe leads to a duct that widens to
With the fifth edition of the American College of form a lactiferous sinus under the nipple–areolar
Radiology BI-RADS in 2013 came some important complex, which then exits the nipple. The fibrous
changes to the breast MR imaging lexicon as well stroma, commonly referred to as Cooper’s liga-
as the mammography and breast ultrasound im- ments, consists of bands of connective tissue
aging lexicons. Several of these changes worked that traverse the breast and insert into the dermis.
to unify terms across breast imaging modalities The breast parenchyma is enclosed by the super-
and eliminate confusing or underused terms.3 ficial fascia, which lies just beneath the skin, and
the deep fascia, which envelops the pectoralis ma-
ANATOMY jor muscle. Overlying the superficial fascia is the
Knowledge of the breast anatomy can assist skin.4 Fig. 1 illustrates these basic anatomic struc-
with providing helpful descriptions and properly tures of the breast as seen on a typical sagittal MR
relaying important information, particularly for imaging sequence.
mri.theclinics.com

Disclosure Statement: The author has nothing to disclose.

Department of Diagnostic Radiology, Breast Imaging Section, The University of Texas MD Anderson Cancer
Center, 1515 Holcombe Boulevard, Unit 1350, Houston, TX 77030, USA
E-mail address: [email protected]

Magn Reson Imaging Clin N Am - (2017) -–-

https://doi.org/10.1016/j.mric.2017.12.001
1064-9689/17/Ó 2017 Elsevier Inc. All rights reserved.
2 Martaindale

long used for mammography, is new for this edi-

tion of the BI-RADS lexicon. On MR imaging,
the amount of fibroglandular tissue should be
assessed on a T1-weighted image using 1 of 4 de-
scriptors: almost entirely fat, scattered fibrogland-
ular tissue, heterogeneous fibroglandular tissue, or
extreme fibroglandular tissue (Fig. 2).
Also new is the addition of a statement on the
BPE, which is defined based on the enhancement
of the patient’s fibroglandular tissue on the first
postcontrast sequence. Descriptor categories
Fig. 1. Sagittal delayed postcontrast image demon-
include minimal, mild, moderate, or marked.
strating major anatomic components of the breast.
Important to note is that the term “multiple foci”
has been removed from the lexicon with the under-
THE MR IMAGING BREAST IMAGING standing that this represents a pattern of back-
REPORTING AND DATA SYSTEMS LEXICON ground enhancement. The distinction between
the amount of fibroglandular tissue and BPE is
The lexicon is divided into sections dedicated to important because the degree of BPE is not
detailing each component of a breast MR imaging necessarily related to the amount of breast tissue
examination. These sections begin with general (Fig. 3). The degree of BPE is largely influenced
breast parenchyma characteristics, such as the by hormonal factors related to the menstrual cycle
amount of fibroglandular tissue and the degree and it is generally recommended that nonurgent
of background parenchymal enhancement (BPE). (screening or follow-up) breast MR imaging be
Next, we review multiple types of lesion findings, scheduled during days 7 to 15 (week 2) of the men-
including focus, mass, nonmass enhancement, strual cycle to minimize these effects.5 Although
multiple benign entities, associated findings previously theorized that BPE may limit the ability
of importance when evaluating a malignancy, of MR imaging to adequately detect malignancy,
enhancement kinetics, and, last, implant evaluation. that has not been found to be the case. Studies
evaluating the effects of BPE on the accuracy of
BREAST TISSUE breast MR imaging have shown that an increased
amount of BPE may lead to an increased recall
The addition of a descriptor of the amount of fibro- rate, but does not decrease the cancer detection
glandular tissue within the breast, similar to that rate.6,7

Fig. 2. Axial precontrast T1-weighted images show examples of almost entirely fat (A), scattered fibroglandular
tissue (B), heterogeneous fibroglandular tissue (C), and extreme fibroglandular tissue (D).
 Breast MR Imaging 3

Fig. 3. Axial precontrast T1-weighted (A) and subtraction image (B) of the same patient at the same level demon-
strating heterogeneous fibroglandular tissue with minimal background parenchymal enhancement.

Included in the discussion of BPE is a descriptor Focus

for symmetric versus asymmetric enhancement
What is the difference between a focus and a
(Fig. 4). Symmetric enhancement is benign bilateral,
mass? According to the American College of Radi-
“mirror image” enhancement that often appears in
ology BI-RADS lexicon, a focus is punctate
the upper outer quadrants and the inferior breasts.
enhancement that is too small to characterize
In prior versions of the BI-RADS lexicon, this was
and has no precontrast correlate (Fig. 5). In prior
referred to as “sheets of enhancement,” a term
versions, a size criterion of less than 5 mm was
now removed from the lexicon. Asymmetric
set, but the current version sets no strict size
enhancement may be benign or malignant and de-
criteria because some masses are now identifiable
scribes enhancement that is more prominent in
and describable at this size with improved MR im-
1 breast. Clinical history of prior radiation therapy
aging technology resulting in better resolution.
for breast cancer can explain asymmetric BPE,
Now the difference between a focus and a mass
because the irradiated breast will have less
is determined by the interpreting radiologist. Fea-
enhancement than the nonirradiated breast.
tures that may assist with differentiating a benign
focus from a malignant one are compared in
BREAST LESIONS Table 1. Most foci are benign, related to hormonal
Next, we discuss the updated terms for findings stimulation or underlying benign lesions (fibro-
within the breast. Many of the terminology cystic changes, fibroadenoma, cyst), but can
changes in this section were enacted to bring rarely represent an early malignancy.8,9 A recent
more consistency in reporting between breast study specifically assessed how frequently foci
MR imaging and other modalities. This update identified on MR imaging were malignant using
aids in clear communication of findings between follow-up MR imaging over a 5-year period and
radiologists as well as the referring clinicians. found that about 97% of foci identified were stable
Applying the appropriate descriptors can also or disappeared on follow-up, and were therefore
assist the radiologist with decision making for benign. The malignant foci were identified by bi-
each lesion, because certain terms carry a higher opsy after an increase in size on MR imaging
risk of malignancy and may guide toward a biopsy and/or suspicious change in characteristics of
or follow-up. the focus in only 2.9% of cases.10

Fig. 4. Axial subtraction images demonstrating symmetric background parenchymal enhancement (BPE) (A) and
asymmetric BPE (B).
4 Martaindale

Fig. 5. Example of a focus denoted by a red arrow on axial subtraction image (A) with no correlate on axial pre-
contrast T1-weighted image (B).

Mass enhancement include homogeneous, heteroge-

neous, rim enhancement, and dark internal septa-
Contrasting with a focus, a mass is large enough to
tions (Fig. 8). Rim enhancement is usually a
be described and is a space-occupying lesion that
malignant feature compared with other enhance-
may present with a variety of shapes, margins, and
ment descriptors,11–13 but cysts and fat necrosis
internal enhancement characteristics. Terms avail-
should first be excluded using other sequences
able to describe the shape include oval, round,
and/or available mammograms. The presence of
and irregular (Fig. 6). Of note, the term oval now
dark internal septations suggests a fibroadenoma
encompasses the former term lobulated, which
if other features support benignity. The terms
was removed from the current lexicon. Margins
“central enhancement” and “enhancing internal
may be described as circumscribed (which was
septations” have been removed owing to a lack
previously called “smooth”), irregular, or spicu-
of clinical relevance.
lated (Fig. 7). To describe a mass as having cir-
cumscribed margins, the entire mass must be
circumscribed. If any part of the mass has irregular Nonmass Enhancement
or spiculated margin, that descriptor should be Our next finding for discussion is nonmass
used, because those terms convey a higher level enhancement (previously called nonmass-like
of suspicion. Studies evaluating the positive pre- enhancement). This describes enhancement that
dictive value of lesion characteristics on MR imag- is discrete from the BPE but is neither a focus
ing have found a spiculated margin to have the nor a mass. One important component of the
highest likelihood of malignancy compared with description of nonmass enhancement is the distri-
other shape and margin descriptors.11–13 Although bution. Options to describe the distribution include
the term indistinct is not included in this version of focal, linear, segmental, regional, multiple regions,
the lexicon, the term is under consideration for an and diffuse (Fig. 9). Focal distribution is enhance-
upcoming edition. Descriptors of internal mass ment within a single duct and involving less than a
quadrant. Linear distribution is within a single or
branching line, and includes enhancement previ-
Table 1 ously described as “ductal,” which has been
Features of a focus eliminated from the lexicon. Segmental distribu-
tion is triangular or conical-shaped, with the apex
Malignant Features Benign Features oriented toward the nipple. Regional distribution
Not bright on Bright on describes enhancement, which involves at least
T2-weighted T2-weighted imaging a quadrant, is geographic, and has no convex
imaging Fatty hilum margins. Multiple regions describe the same
No fatty hilum Persistent kinetics appearance as regional distribution but at least 2
Washout kinetics Stable since prior broad areas separated by normal tissue. Diffuse
Larger or new since examination
distribution is widely scattered and evenly distrib-
prior examination
uted throughout the tissue. Diffuse distribution is
 Breast MR Imaging 5

Fig. 6. (A) Sagittal delayed postcontrast fat saturated T1-weighted image demonstrates an oval mass. Axial sub-
traction images demonstrate a round mass (B) and an irregular mass (C).

generally benign whereas segmental or linear dis- inclusion of this section largely serves to assist
tribution has a higher risk of malignancy.11,14–17 readers in identifying and separating these rela-
The next component of nonmass enhancement tively common findings from ones that require
description is the pattern of enhancement. Non- additional evaluation or biopsy. Specifically
mass enhancement patterns of enhancement discussed are intramammary lymph nodes, skin
include homogeneous, heterogeneous, clumped, lesions, and nonenhancing findings. Intramam-
and clustered ring (Fig. 10). Clumped enhance- mary lymph nodes are a common benign
ment is a suspicious finding11,14,15 and has previ- mammographic finding, so it should come as
ously been described as a cobblestone or string no surprise that they would be commonly identi-
of pearls appearance. Clustered ring enhance- fied on MR imaging as well. Intramammary lymph
ment is a new term and describes multiple thin nodes are circumscribed, reniform, homoge-
rings of enhancement. Clustered ring enhance- neously enhancing masses with hilar fat; are usu-
ment is a suspicious finding, although the differen- ally less than 1 cm in size; and are most
tial diagnosis includes benign entities such as duct commonly located in the upper outer quadrant.
ectasia, fibrocystic change, and papilloma in addi- A close association with a vessel is a helpful
tion to malignancy (ductal carcinoma in situ and feature. Skin lesions such as keloids, sebaceous
invasive carcinoma). Studies have identified a pos- cysts, postsurgical scar, and dermatitis can also
itive predictive value for malignancy with clustered be identified as enhancing findings within the
ring enhancement from 77% to 100%.15–17 Of skin, and are benign. Several nonenhancing find-
note, the terms “reticular” and “dendritic” have ings are also described (Fig. 11). Duct ectasia
been eliminated from the lexicon owing to can be identified by the high signal intensity
underuse. within the ducts on precontrast T1-weighted im-
ages. Cysts may appear as signal voids on sub-
traction images, but will have associated high
Benign and Nonenhancing Findings
intensity T2-weighted signal. Varying degrees of
Also new for this edition is a section on special susceptibility artifact can be seen in association
topics. These entities are all benign, and the with biopsy marker clips and other metallic

Fig. 7. Axial subtraction image demonstrates a mass with circumscribed margin, denoted by a red arrow (A).
Sagittal postcontrast fat-saturated T1-weighted images demonstrate mass with irregular margin (B) and a mass
with spiculated margin (C).
6 Martaindale

Fig. 8. Axial subtraction image of an oval mass with homogeneous enhancement (A). Sagittal postcontrast fat-
saturated T1-weighted image of an irregular mass with heterogeneous enhancement (B). Axial subtraction im-
ages of an oval mass with dark internal septations (C) and a round mass with rim enhancement denoted by
the red arrow (D).

foreign bodies, and correlation with mammog- patients and can alter staging and surgical man-
raphy can assist in identifying the source of agement in those patients who have already
susceptibility artifacts. Knowing the patient’s been diagnosed with breast cancer. Just as on
treatment history is important in correctly identi- mammography, nipple retraction can be identified
fying skin and trabecular thickening as benign on MR imaging. Nipple invasion by a mass is an
postradiation changes. Clinical history can also important finding and can alter the surgical options
assist with correctly identifying nonenhancing available to the patient. Direct skin invasion and
architectural distortion from postbiopsy or post- pectoralis muscle or chest wall invasion should
surgical changes as well as postoperative collec- be described if present to allow for appropriate
tions. Sequences obtained without and with fat treatment planning. Sometimes these findings
suppression can assist in accurately identifying can be quite subtle on other imaging modalities
benign, fat-containing lesions such as lymph or not recognized at all without the use of MR im-
nodes, fat necrosis, hamartoma (Fig. 12), and aging. Although we have seen examples of benign
postoperative collections with fat. skin thickening associated with postradiation
changes, skin thickening can also be seen as an
associated feature of breast malignancy. An
Associated Features
important differentiating factor is concurrent
There are several associated features that can be enhancement in the area of skin thickening. Suspi-
identified on MR imaging examinations that in- cious axillary lymphadenopathy will demonstrate
crease suspicion for breast cancer in undiagnosed effacement of the fatty hilum and a heterogeneous
 Breast MR Imaging 7

Fig. 9. Axial subtraction images demonstrate focal nonmass enhancement (NME) (A), linear NME (B), segmental
NME (C), and regional NME (D). Axially oriented maximum intensity projection image demonstrates a mass and
diffuse NME (E).

enhancement pattern compared with the homoge- KINETIC CURVE ASSESSMENT

neous pattern seen with benign/normal lymph
nodes. Fig. 13 provides examples of several of Once the breast lesions of interest have been
these associated features. identified, enhancement kinetics characteristics
should be evaluated and reported for each finding.

Fig. 10. Axial subtraction image shows example of clumped nonmass enhancement (NME) (A). Sagittal delayed
postcontrast T1-weighted image demonstrates clustered ring NME, annotated by the red arrows (B).
8 Martaindale

Fig. 11. Nonenhancing findings. Axial precontrast T1-weighted image with red arrow denotating the high signal
intensity within multiple ducts, consistent with duct ectasia (A). Axial subtraction image and fat-saturated
T2-weighted images from the same patient demonstrate oval nonenhancing signal voids (B) with high signal in-
tensity (C) on T2-weighted images, consistent with cysts. Sagittal delayed postcontrast fat-saturated image with
susceptibility artifact from biopsy marker clip (D). Axial postcontrast fat-saturated image demonstrating nonen-
hancing left breast skin thickening and trabecular thickening in a patient with history of left breast cancer and
treatment changes from radiation therapy (E).

Fig. 12. Right mediolateral oblique mammogram (A), sagittal postcontrast fat-saturated T1-weighted image (B)
and axial precontrast non–fat-saturated image (C) of an oval fat containing mass, consistent with hamartoma.
 Breast MR Imaging 9

Fig. 13. Associated features. Axial subtraction image demonstrates nipple invasion by malignancy (A). Axial post-
contrast fat-saturated image with diffuse skin and trabecular thickening (B) that enhances, making this a suspi-
cious feature compared with the nonenhancing skin and trabecular thickening associated with radiation therapy
shown in Fig. 11. Sagittal postcontrast fat saturated image with direct skin invasion by the breast malignancy (C).
Axial subtraction image demonstrating pectoralis major and skin invasion by the malignancy (D). Axial subtrac-
tion image showing matted axillary lymphadenopathy in a patient with known breast cancer (E).

Enhancement kinetics are evaluated in 2 phases, kinetics,12,14,17 others have not found kinetics to
the initial phase and the delayed phase (Fig. 14). be significant predictors of malignancy.11,16 It is
The initial phase is within the first 2 minutes and for this reason that the morphology of the lesion
is described as slow, medium, or fast. Slow initial should be the most important factor for deciding
phase enhancement is a less than 50% increase
in signal intensity. Medium initial phase enhance-
ment is a 50% to 100% increase, and fast initial
phase enhancement is a greater than 100%
increase in signal intensity. The delayed phase
is described as persistent, plateau, or wash-
out. Persistent enhancement is continuously
increasing. Plateau enhancement demonstrates
no change after the initial phase, and washout
shows decreasing signal intensity.
The most suspicious kinetic feature should be
reported, because any one lesion may display
many different types of kinetic enhancement.
Although some studies have shown malignant le-
sions are more likely to demonstrate washout Fig. 14. Kinetic curve assessment.
10 Martaindale

Fig. 15. Signs of intracapsular implant rupture. Sagittal silicone-specific sequences demonstrate the keyhole sign
(A) and the linguine sign (B), annotated by the red arrows. Note the white ball at the end of the keyhole sign
compared with the black ball seen on the radial fold in Fig. 16.

if a lesion requires additional evaluation or biopsy. fold has a dark ball at the end of the fold, whereas
For example, an intramammary lymph node may the keyhole sign has a white ball at the end owing
display a washout pattern, but the morphology is to the silicone that is outside of the implant settling
classic for a common, benign entity that should between the layers of the collapsing implant wall.
not be referred for additional evaluation or biopsy. Extracapsular implant rupture occurs when both
the implant wall and the fibrous capsule have
IMPLANT ASSESSMENT broken, and silicone is extruding into the adjacent
breast tissue (Fig. 17).18,19 With silicone-sensitive
The section on implant assessment is also new for sequences, extracapsular silicone can also
this edition of BI-RADS. Implant evaluation cannot be detected in the axillary lymph nodes. Also
be adequately performed on a standard breast sometimes seen is a periimplant fluid collection,
MR imaging because a dedicated protocol consist- which can be present in recent implantation,
ing of water and silicone-specific sequences is
required, and images in 2 planes should be
reviewed, because single plane “ruptures” may
simply be a fold in the implant.18 Evaluation of im-
plants includes a description of lumen type (single
lumen vs double lumen), location (prepectoral or
retropectoral), and any findings to suggest rupture
or signs of implant complications. Intracapsular
rupture occurs when the implant ruptures, but the
fibrous capsule formed by the body around the
implant remains intact. The intact fibrous capsule
keeps the silicone from extending into the adjacent
breast tissue, and the implant shape may seem to
be normal mammographically. MR imaging signs
of intracapsular rupture include the subcapsular
line, keyhole sign, and the linguine sign (Fig. 15).
It important to differentiate these signs of intracap- Fig. 16. Radial fold. Axial silicone-specific sequence
sular rupture from the normal undulations of the shows a dark line with a black ball at the end along
implant, also known as radial folds (Fig. 16). A radial the medial aspect of the implant.
 Breast MR Imaging 11

Fig. 17. Signs of extracapsular rupture, annotated by red arrows. Axial silicone-specific sequence (A) shows ex-
tracapsular extrusion of silicone along the chest wall. Also note the subcapsular line along the anterior portion
of the implant. Sagittal silicone-specific sequence (B) shows extracapsular extrusion of silicone along the superior
aspect of the implant.

bleeding, and infection (Fig. 18). Although very rare, identified a higher incidence in patients with
it is important to be aware of the occurrence of textured surface implants. The most common pre-
breast implant–associated anaplastic large cell sentation is a persistent seroma with delayed onset
lymphoma. In March 2017, the US Food from the time of implantation, although some pa-
and Drug Administration released a report detail- tients did present with a mass.20,21
ing 359 medical device reports of breast
implant–associated anaplastic large cell lym- FINAL ASSESSMENT
phoma, including 9 deaths.20 A recent article by Sri-
nivasa and colleagues21 reviewed reports from 40 Just as with mammography and ultrasound exam-
countries and identified 340 unique cases of breast inations, a final BI-RADS assessment category
implant–associated anaplastic large cell lym- should be assigned for each study. The final
phoma with 5 associated deaths. Both groups assessment categories are the same for MR

Fig. 18. Axial silicone only sequence (A) showing large periimplant fluid collection, no silicone rupture, and the
pattern from the textured implants visible along the edges of the implant. Subtraction image (B) showing no
enhancement within the fluid collection. The fluid was aspirated for preoperative cytology and microbiology
studies revealing no evidence of infection or malignancy.
12 Martaindale

imaging, as for mammography and ultrasound short-interval follow-up. AJR Am J Roentgenol

examination: 2011;196:218–24.
7. DeMartini WB, Liu F, Peacock S, et al. Background
Category 0: Incomplete, need additional imag- parenchymal enhancement on breast MRI: impact
ing evaluation on diagnostic performance. AJR Am J Roentgenol
Category 1: Negative 2012;198:W373–80.
Category 2: Benign 8. Ha R, Sung J, Lee C, et al. Characteristics and outcome
Category 3: Probably benign of enhancing foci followed on breast MRI with man-
Category 4: Suspicious agement implications. Clin Radiol 2014;69:715–20.
Category 5: Highly suspicious of malignancy 9. Liberman L, Mason G, Morris EA, et al. Does size
Category 6: Known biopsy-proven malignancy matter? Positive predictive value of MRI-detected
Although not currently mandated by the US Food breast lesions as a function of lesion size. AJR Am
and Drug Administration, assigning a final BI-RADS J Roentgenol 2006;186:426–30.
code to each examination is encouraged to aid in 10. Clauser P, Cassano E, De Nicolo A, et al. Foci on
communication and provide continuity between breast magnetic resonance imaging in high-risk
mammogram, ultrasound, and MR imaging reports. women: cancer or not? Radiol Med 2016;121:611–7.
Of note, implant assessment MR imaging studies 11. Liberman L, Morris E, Lee M, et al. Breast lesions de-
should not have a final BI-RADS assessment. tected on MR imaging: features and positive predic-
tive value. AJR Am J Roentgenol 2002;179:171–8.
12. Baltzer P, Benndorf M, Dietzel M, et al. False-positive
SUMMARY findings at contrast-enhanced breast MRI: a
The updated breast MR imaging BI-RADS lexicon BI-RADS descriptor study. AJR Am J Roentgenol
serves to unite terminology and recommendations 2010;194:1658–63.
across multiple modalities, improving communica- 13. Nunes L, Schnall M, Orel S. Update of breast MR im-
tion between radiologists and referring providers. aging architectural interpretation model. Radiology
The unification of terminology eliminated confusing 2001;219:484–94.
or underused descriptors in favor of terms recog- 14. Gity M, Moghadam K, Jalali A, et al. Association of
nized from other, more established modality lexi- different MRI BIRADS descriptors with malignancy
cons. Consistency in reporting improves patient in non mass-like breast lesions. Iran Red Crescent
care by providing guidelines for lesion management. Med J 2014;16(12):e26040.
15. Tozaki M, Fukuda K. High-spatial-resolution MRI of
non-masslike breast lesions: interpretation model
REFERENCES based on BI-RADS MRI descriptors. AJR Am J
Roentgenol 2006;187:330–7.
1. American College of Radiology. Breast Imaging
16. Uematsu T, Kasami M. High-spatial-resolution 3T
Reporting and Data System (BI-RADS). 1st edition.
breast MRI on nonmasslike enhancement lesion: an
Reston (VA): American College of Radiology; 1992.
analysis of their features as significant predictors of
2. American College of Radiology. Breast Imaging
malignancy. AJR Am J Roentgenol 2012;198:1223–30.
Reporting and Data System (BI-RADS). 4th edition.
17. Tozaki M, Igarashi T, Fukuda K. Breast MRI using the
Reston (VA): American College of Radiology; 2003.
VIBE sequence: clustered ring enhancement in the
3. Morris EA, Comstock CE, Lee CH, et al. ACR differential diagnosis of the lesions showing non-
BI-RADSÒ magnetic resonance imaging. In: ACR masslike enhancement. AJR Am J Roentgenol
BI-RADSÒ atlas, Breast Imaging Reporting and 2006;187:313–21.
Data System. Reston (VA): American College of 18. Shah M, Tanna N, Margolies L. Magnetic resonance
Radiology; 2013. imaging of breast implants. Top Magn Reson Imag-
4. McGuire KP. Breast anatomy and physiology. In: ing 2014;23:345–53.
Aydiner A, Igci A, Soran A, editors. Breast disease: 19. Gorczyca D, Gorczyca S, Gorczyca K. The diag-
diagnosis and pathology. New York: Springer Inter- nosis of silicone breast implant rupture. Plast Re-
national Publishing; 2016. p. 1–14. constr Surg 2007;120(7 Suppl 1):49S–61S.
5. Giess CS, Yeh ED, Raza S, et al. Background paren- 20. US Food and Drug Administration. Breast
chymal enhancement at breast MR imaging: normal implant-associated anaplastic large cell lymphoma
patterns, diagnostic challenges, and potential for (BIA-ALCL). 2017. Available at: www.fda.gov. Ac-
false-positive and false-negative interpretation. Ra- cessed March 22, 2017.
diographics 2014;34:234–47. 21. Srinivasa D, Miranda R, Kaura A, et al. Global adverse
6. Hambly NM, Liberman L, Dershaw DD, et al. Back- event reports of breast implant-associated ALCL: an
ground parenchymal enhancement on baseline international review of 40 government authority data-
screening breast MRI: impact on biopsy rate and bases. Plast Reconstr Surg 2017;139:1029–39.