Toward a Standardized Breast Ultrasound Lexicon BI-RADS Ultrasound

Ellen B Mendelson Wendie A Berg and Christopher R B Merritt

SONOGRAPHY IS an imp0l1ant complement to mammography Analysis of sonographic

features may aid in appropriate selection of lesions for follow-up or biopsy Recognizing that consisshytent terminology and clear communication of findshyings and results directly affect patient manageshyment a validated lexicon of descriptors for breast sonography is in development This includes terms to describe shape orientation margin regularity and thickness matrix echogenicity matrix homoshygeneity acoustic attenuation and effect on surshyrounding ti ssue

With support from the Offlce on Womens Health Department of Health and Human Sershyvices the American College of Radiology (ACR) Commission on Ultrasound has developed the initial draft of a breast ultrasound lexicon to be used for standardized reporting of breast ultrashysound findings In tended to complement the ACR Breast Imaging Reporting and Data System (BIshyRADSTM) lexicon the breast ultrasound lexicon (BI-RADS Ultrasound) will undergo final review in mid 2001 and is expected to be released in December 2001 An overview of the proposed lexicon and examples of features described in the lexicon are presented

INTRODUCTION

Although mammography remains the most senshysitive method for detecting preclinical breast carshycinoma its limited specificity resu lts in need to biopsy many abnormalities to determine whether they are benign or malignant I 2 Indi cations for breast sonography include the following the in itial eva luation of palpable abnormalities in women under 30 initial identification and characterization of palpable and nonpalpable abnormalities guidshyance of interventional procedures and evaluation of problems associated with breast implants34

Several recent studies suggest that sonography in combination with mammography can reduce the number of false-positive recommendations for bishyopsyS-B Mammography remains the standard for breast screening as most ductal carc inoma in situ is missed sonographically9-1 I

The growing use of ultrasonography has created the need for a standardized method for lesion characterization descri ption and reporting1 2 The

mammography lexicon developed by the ACR the Breast Imaging Reporting and Data System (BI-RADSTM)13 provides sta ndardized assessment and associated management recommendations for masses and calcifications Based on success of BI-RADS with mammography the development of a lexicon for breast ultrasound (BIRADS Cltrashysound) and breast magnetic resonance imaging 14 has been a high priority for the ACR The lexicons are designed to use shared terminology whenever possible When com pleted the lexicons will aid refelTing physician s radiologists and patients in describing abnormalities and understanding their management implications Furthermore these lexshyicons will provide a basis for validation of outshycomes across multiple centers

A breast ultrasound lexicon_ the Breast Imaging Reporting and Data System lJtrasound is curshyrently being developed by the ACR The initial draft was prepared by the Breast Cltrasound Lexshyicon Subcommittee of an Expert Working Group to Plan and Develop Protocol s for Optimization and Clinical Testing of Breast Ultrasound supported by a contract from the Offlce of Womens Health National Institutes of Health and conducted by the Commission on Ultrasound of the ACR Techshyniques adapted from those used in the development of BI-RADS are being used in the formulation of the new ultrasound lexicon The ACR lexicon is ex pected to be completed and released j n late 200 1 follo wing its validation

Froll the Deporll11enl of Radiology Western Pennsylvania

Hospital PitTsiJurgh PA Depamnenr of Diagnostic Radiology

and Greenebawll Cancer Center UniversiT) of Maryland Baltimore MD alld The DeparTment of Radiology Thomas

j efferson UniversilY j efferson Medical College Philadelphia

PA This work Ivas supPOrTed in part by COnTract 282-97-0076

Federal Technology Tran sfer Program to Advance Novel

BreasT Imaging Technologies 15 Public HealTh Service Office

on Womens Health CS D epaJ1meJl of HealTh and Human

Senices Address reprinT requests To Ellen B Mendelson MD Deshy

palllnelll of Radiology The WeSTern Penl1Sylvania Hospital 4800 Friendship AFe PillsiJlrgh PA 15224

Copyright copy 2001 by WB Salnders Company

0037- 98x013603-0007$3500O

doi JOI053sroe200125125

Seminars in Roentgenologv Vol XXXVI No 3 (July) 2001 pp 217-225 217

218 MENDELSON BERG AND MERRITI

Fig 1 Shape Masses are described as round (A) oval (B) or irregular (C) A round mass is spherical ball-shaped or globular in shape An oval mass is elliptical or egg-shaped These descriptors are similar to those used in the ACR BI-RADS lexicon for mammography (A) A spherical simple cyst is shown (B) An oval fibroadenoma is shown The invasive cancer shown in C is irregular

FEATURE ANALYSIS

Lesion features include primary fea tures such as shape (Fig 1) orientation (Fig 2) margins (Figs 3 through 5) matri x echogenicity and homogeneity (Figs 6 and 7) and attenuation (Fig 8) which should be described and applied in a consistent fashion In addition secondary associated findings such as architectural distortion (Fig 9) retraction or angulation of Coopers ligaments (Fig 9B) di lated ducts calcifications (Fig 10) and changes in the ski n subcutaneous fat and pectoral muscle can be recorded as well These fea tures of masses have been enumerated previously5715 16 The utility of

each category of features requires validati on along with that of individual descriptors For example matrix homogeneity probably has less specificity and significance than description of mass marg ins The most appropriate descriptor for each category of characteristics shou ld be applied whe n desc ribshying a lesion (Table 1) Wherever poss ible fea ture descriptors similar to those used by BI-RADS fo r mammography have been imported As with mamshymography use of the lexicon is predicated on

Fig 2 Orientation Mass orientation refers to the relationshyship of the long axis of the mass to the skin This may be parallel as is common with fibroadenomas (A) or not parallel a common orientation for many cancers (B) Because round lesions do not have one axis that is longer than another they should be classified as not parallel

Fig 3 Margins The margin is the boundary between the lesion and its surroundings Several characteristics of the lesion margin are important The margin should be described as circumscribed (distinct and smooth) (A) or irregular (indisshytinct microlobulated angUlar or spiculated) (B) In A the circumscribed margin of a fibroadenoma is shown In B an invasive cancer demonstrates an ill -defined or indistinct inshyterface with the surrounding tissues As in mammography ill-defined margins are associated with higher risk of maligshynancy than circumscribed margins

exce llent sonographic technique using a linear transd ucer whose center frequency is at least 7 MHz Documentation should be performed in acshycord with the American Colleoe of Radioloayeo c

Standards Orthogonal v iews of the lesion should be obtained and the orientation of the transducer and location of the abnormality should be deshyscribed using quadran t clock-face location andlor labeled diagram of the breast ideall y including distance from the nipple

Several previous studies5 15 reach the conclushysion that multiple features must be analyzed to achie ve as great a specificity as possible in sonoshygraphic charac terizati on As an example the diagshynosis of a mass as a si mple cyst requires that the shape be rou nd oval or gently lobul ated margin circumscribed echogenicity (echo pattern) aneshychoic and that there be acoustic enhancement Based on these combined features the impress ion is that of a simple cyst (Figs 1 5 and 6) The final assessme nt for the combination of mammographic

Fig 4 Margin thickness Circumscribed margins may be thin thick or of mixed thickness The fibroadenoma (A) has a thin and distinct boundary delineating it from surrounding breast tissue The invasive cancer in 4B has generally thick margins whereas the cancer in C shows a margin composed of thick thin and indistinct segments Barely perceptible margins favor benign etiology

218 MENDELSON BERG AND MERRITI

Fig 1 Shape Masses are described as round (A) oval (B) or irregular (C) A round mass is spherical ball-shaped or globular in shape An oval mass is elliptical or egg-shaped These descriptors are similar to those used in the ACR BI-RADS lexicon for mammography (A) A spherical simple cyst is shown (B) An oval fibroadenoma is shown The invasive cancer shown in C is irregular

FEATURE ANALYSIS

Lesion features include primary fea tures such as shape (Fig 1) orientation (Fig 2) margins (Figs 3 through 5) matri x echogenicity and homogeneity (Figs 6 and 7) and attenuation (Fig 8) which should be described and applied in a consistent fashion In addition secondary associated findings such as architectural distortion (Fig 9) retraction or angulation of Coopers ligaments (Fig 9B) di lated ducts calcifications (Fig 10) and changes in the ski n subcutaneous fat and pectoral muscle can be recorded as well These fea tures of masses have been enumerated previously5715 16 The utility of

each category of features requires validati on along with that of individual descriptors For example matrix homogeneity probably has less specificity and significance than description of mass marg ins The most appropriate descriptor for each category of characteristics shou ld be applied whe n desc ribshying a lesion (Table 1) Wherever poss ible fea ture descriptors similar to those used by BI-RADS fo r mammography have been imported As with mamshymography use of the lexicon is predicated on

Fig 2 Orientation Mass orientation refers to the relationshyship of the long axis of the mass to the skin This may be parallel as is common with fibroadenomas (A) or not parallel a common orientation for many cancers (B) Because round lesions do not have one axis that is longer than another they should be classified as not parallel

Fig 3 Margins The margin is the boundary between the lesion and its surroundings Several characteristics of the lesion margin are important The margin should be described as circumscribed (distinct and smooth) (A) or irregular (indisshytinct microlobulated angUlar or spiculated) (B) In A the circumscribed margin of a fibroadenoma is shown In B an invasive cancer demonstrates an ill -defined or indistinct inshyterface with the surrounding tissues As in mammography ill-defined margins are associated with higher risk of maligshynancy than circumscribed margins

exce llent sonographic technique using a linear transd ucer whose center frequency is at least 7 MHz Documentation should be performed in acshycord with the American Colleoe of Radioloayeo c

Standards Orthogonal v iews of the lesion should be obtained and the orientation of the transducer and location of the abnormality should be deshyscribed using quadran t clock-face location andlor labeled diagram of the breast ideall y including distance from the nipple

Several previous studies5 15 reach the conclushysion that multiple features must be analyzed to achie ve as great a specificity as possible in sonoshygraphic charac terizati on As an example the diagshynosis of a mass as a si mple cyst requires that the shape be rou nd oval or gently lobul ated margin circumscribed echogenicity (echo pattern) aneshychoic and that there be acoustic enhancement Based on these combined features the impress ion is that of a simple cyst (Figs 1 5 and 6) The final assessme nt for the combination of mammographic

Fig 4 Margin thickness Circumscribed margins may be thin thick or of mixed thickness The fibroadenoma (A) has a thin and distinct boundary delineating it from surrounding breast tissue The invasive cancer in 4B has generally thick margins whereas the cancer in C shows a margin composed of thick thin and indistinct segments Barely perceptible margins favor benign etiology

TOWARD A STANDARDIZED BREAST ULTRASOUND LEXICON 21 9

Fig 5 Margins The smooth barely perceptible margin of a simple cyst (A) is contrasted with the irregular margins of a small invasive cancer (B) Margin irregularities may include angular edges m icrolobulation and spiculation

and sonographic fi ndings using BI-RADS is cateshygory 2 benign with routine screening recomshymended The accuracy of sonographi c identificashyti on of cysts approaches 100 provided stric t ad herence to the classical sonographic charac teri sshytics are observed 17

t

Practical use of any lexicon req uires an undershystanding of the de fin iti ons of each term An examshypJe is complicated versus complex The presence of homogeneous low-level internal echoes throug holl t a cystic lesion that has all the other features of a sim ple cyst as above results in its designation as complicated 18 Many of these masses appear solid a lbeit benign and may be reported as comshyplicated cyst or probabl y ben ign solid lesion Often such lesions are incidentall y found during ul trasound examination peIiormed for other reashysons Recent studies cumulative ly evaluating 567 incidental compli cated cysts identifi ed only one 3-mm in situ ma lignancy (positive predictive value [PPY] 02)1 9-21 Based on these data short-term follow-up appears appropriate (BI-RADS category 3 probabJy ben ign) although further validation is

Fig 6 Echogenicityechotextureecho pattern The matrix of a mass may be homogeneous or heterogeneous Homogeshyneous masses may be anechoic hypoechoic or hyperechoic Cysts are typically anechoic (A) The echogenicity of hyposhyechoic and hyperechoic masses should be compared with the echogenicity of fibroglandular tissue In this example a hamartoma produces a homog eneous mass (arrows) that is hyperechoic compared with fat (B) Invasive cancer (C) is often hypoechoic both to fat and to glandular tissue (same lesion as in Fig 2B)

Fig 7 Echogenicityechotextureecho pattern Heterogeshyneous masses may be solid or complex conta ining a mixture of solid and cystic components The invasive cancer in A is predominantly hypoechoic but contains some areas of inshycreased echogenicity resulting in a heterogeneous echotexshyture Complex masses may be predominantly solid (B) or cystic (C both intracystic papillary carcinomas) and should be regarded as suspicious for cancer in the absence of clinical finding s suggesting an abscess

encouraged Interval enl argement (mammographishycall y or sonograph ical ly) or the presence of any suspicious features should prompt aspi ration and possible core biopsy if it proves solid 22

Suspicious features include an intracystic mass mural nodule thick septations or a thi ck or irregshyul ar wall When such fea tures are present the mass should be desc ribed as a complex cystic mass (Figs 7B and C) These lesions generall y requ ire aspiration or biopsy (BI-RADS category 4 susp ishycious)

When a so lid lesion is presen t careful analysis of contour margins matri x and attenuation may a llow classification of some nodules as BI-RADS category 3 probab ly benign and provide the opshyti on of short interva l follow-up at 6 months 12 months and 24 mon ths rather than biopsy5-7 As in ma mmography for a lesion to be assessed as probably benign it should have lt2 risk of malignancy 2324 As mentioned preliminary data suggest cysts with internal echoes can be so classhysified as can clusters of tin y cystic foci wi th thin intervening septatio ns compatible with apocrine metaplas ia 25

Fig 8 ShadOWingenhancement Enhancement may be seen with cysts (A) fibroadenomas (B) and high-grade inshyvasive cancers (C) Central shadowing is associated with the small invasive cancer is shown in C Refractive edge shadowing (present in A and B) is excluded from considershyation A single echogenic calcification is present in the fibroshyadenoma (8)

220

Fig 9 Effect on surrounding tissue Masses may affect surrounding breast tissues resulting in architectural distorshytion (A) Extensive distortion of breast architecture by a small scirrhous cancer is shown (8) A more subtle change is seen where a Cooper ligament is interrupted by a cancer (open arrows)

Stavros et a]5 proposed three categories of solid lesions that could be class ified as BI-RADS cateshygory 3 probably benign in the absence of any suspicious features (I) masses with intense and uniform hyperechogenicity re lative to fat (2) masses with an ellipsoid shape and thin echogenic capsule and (3) masses with two or three gentle lobulations and a thin echogenic capsu le Individshyually each of these characteristics had a negative predictive va lue for malignancy of 988 to 100905 Although accepted by some it is important to note that these results have not been validated across multiple centers Indeed one recent study26

suggests that not all readers achieve sufficiently high specificity to fo llow so lid lesions [n the draft BI-RADSUltrasound lexicon the concept of a thin echogenic capsule or pseudocapsule- has been replaced by that of a thin smooth margin analoshygous to a circumscribed mass mammographishycally Although palpable solid lesions are genershyally recommended for biopsy 2l24 it is not yet clear whether any size criterion or palpability of the lesion influences the absolute risk of malignancy

For solid masses irregularity of shape and margins dominate other features suggesting maligshynancy with a PPY of ma lignancy of 86 to 9357 such lesions are appropriately classified as BI-RADS category 4 or 5 with biopsy recomshymended Other features have lower specificity Orientation of the long axis of the mass nonparalshylel to the skin synonymously termed taller than wide5 has been associated with a 62 to 81 Ok likelihood of malignancy57 and is more common ly seen in cancers lt 1 cm in size Ylost fibroadenoshymas as well as many cancers are oriented with their long axes parallel to the skin (wider than tall) J I

Echotextureecho pattern appears to be less helpful in differentiating benign fro m malignan t sol id

MENDELSON BERG AND MERRITI

masses I as most masses will be hypoechoic to pare nchyma Acoustic attenuation (shadowing) is suspicious for malignancy but as many as 2 1 of benign les ions will show shadowing7 Similarly acoustic enhancement while common in benign lesions may be present in up to 42 of cancerss

Several typically benign lesions are included as special cases (Fig 11) This includes lymph nodes with a thin circumscribed capsule and centra l echogenic hi lum Foreign bodies are spec ial cases and include siliconomas and free extracapsu lar silicone (Fig 11)27 Description of vascu larity of the lesion is not a required standard (Fig 12) as no reliable distinction has yet been made between

30benign and malignant lesions on this basis 2 8shy

Vasc ularity is described as the same increased_ or decreased relative to surrounding parenchyma

Table I is a worki ng draft of a breast ultrasound lexicon including feature categories and descripshytors Although different in format this draft was based largely on the version developed by the Lexicon Committee of the Expert Working Group It is important to note that these recommendations awa it va lidation and are subject to modification before release of the final draft of the ACR BIRADSlltrasound Descriptors are illustrated in Figures 1 through 12 The illustrations show only a

Fig 10 Associated finding s Dilated ducts skin changes and calcifications associated with breast masses may be seen with ultrasonography and should be described when present (A) Macrocalcifications (~0 5 mm) associated with a fibroad shyenoma are shown (arrowheads) (8) Microcalcifications laquo05 mm) in an invasive ductal carcinoma are shown (arrows) (C) Microcalcifications not in a mass and particularly within a tubular dilated duct (curved arrow) are suspicious and may be seen in ductal carcinoma in situ as in this example

bull TC

0 f-

bull

bull

rJ

Ca

se

Re

vie

we

r o

_

gtC

LA

SS

IFIC

AT

ION

CA

TE

GO

RIE

S amp

TE

RM

S

DE

SC

RIP

TIO

N

Ta

ble

1

Dra

ft A

CR

BI-

RA

DS

Ult

raso

un

d L

exi

con

Cla

ssif

ica

tio

n F

orm

4-4

-01

~

Ma

sse

s

ArI

2sect

secto

ccup

ies

spac

e an

d sh

ould

be

seen

in

two

diffe

rent

pr

ojec

tions

Sha

pe

(sel

ect

one)

0O

val

Rou

nd

Irre

gula

r

ellip

tical

or

egg-

shap

ed (

ma

y in

clud

e tw

o o

r th

ree

undu

latio

lls

ie

lob

ular

)

sphe

rical

ba

ll-sh

aped

ci

rcul

ar

or g

lob

ula

r ne

ither

roun

d n

or o

val

Ori

enta

tion

B P

aral

lel

long

axi

s o

f les

ion

orie

nted

alo

ng s

kin

line

(w

ider

than

tal

l)

(sel

ect

one)

N

ot P

aral

lel

no lo

ng a

xis

or

axis

no

t orie

nted

alo

ng s

kin

line

(ta

ller t

han

wid

e)

Mar

gin

C

ircu

msc

ribe

d (s

elec

t al

l tha

t ap

ply)

Y

ES

(d

yes

che

ck o

ne o

ptio

n be

lOW

)0 8

No

perc

eptib

le r

im o

r th

in r

im

Thi

ck r

im

NO

(if

no

go

to i

rreg

ular

)0

IRR

EG

ULA

R (

sele

ct a

ll th

at

appl

y)

0 In

dist

inct

A

ng

ula

r M

icro

lobu

late

d S

picu

late

d ~

Ech

o P

atte

m

A

nech

oic

Hyp

erec

hoic

C

ompl

ex

Hyp

oech

oic

~ (s

elec

t on

e)

Po

ste

ri or

Aco

ustic

N

o po

ster

ior

acou

stic

fea

ture

s F

eatu

res

Enh

ance

men

t S

hado

win

g C

om

bin

ed p

atte

rn

~ S

urro

undi

ng T

issu

e 0

No

effe

ct

BIden

tifia

ble

effe

ct (

sele

ct a

ll th

at a

pp

ly)

Duc

ts

Coo

pers

liga

men

t ch

an

ge

s

0E

dem

a A

rchi

tect

ural

dis

tort

ion

Ski

n T

hick

enin

g

B S

kin

retr

actio

nflr

regu

lari

ty

Pec

tora

l m

uscl

e se

en

bu

t pl

ane

with

an

teri

or

tissu

e is

un

cle

ar

Sm

ooth

dis

tinct

mar

gin

with

thi

n t

hick

or

no p

erce

ptib

le li

near

rim

circ

umsc

ribe

d m

argi

n w

ithin

lin

ear

rim

circ

umsc

ribe

d m

argi

n w

me

asu

rab

le th

ick

rim gt

1 m

m b

etw

een

lesi

on a

nd s

urro

undi

ng t

issu

e

mar

gins

dem

onst

ratin

g a

com

bina

tion

of f

eatu

res

incl

udin

g at

leas

t 1

of t

hose

list

ed b

elow

poor

ly d

efin

ed m

argi

n pa

rt o

r al

l of t

he m

argi

n fo

rmed

by

shar

p lin

ear

inte

rsec

tions

that

form

acu

te a

ngle

s m

argi

n ch

arac

teri

zed

by b

y gt

3 sm

all

sho

rt c

ycle

und

ulat

ions

m

argi

n ch

arac

teri

zed

by

shar

p pr

ojec

ting

lines

with

out

inte

mal

ech

oe

s de

fined

rel

ativ

e to

fat

eq

ua

l to

fibro

glan

dula

r tis

sue

com

bine

d cy

stic

(an

echo

ic)

and

ech

og

en

ic c

om

po

ne

nts

de

fined

rel

ativ

e to

fib

rogl

andu

lar

tissu

e i

soec

hoic

or

hyp

oe

cho

ic to

fat

con

tain

s lo

w-l

evel

ecl

hoes

th

roug

hout

(eg

co

mpl

icat

ed c

yst o

r fib

road

enom

a)

no p

oste

rior

sha

dow

ing

or e

nh

an

cem

en

t in

crea

sed

post

erio

r ec

hoes

de

crea

sed

post

erio

r ec

hoes

ex

clud

ing

edge

sh

ad

ow

s bo

th s

hado

win

g an

d e

nh

an

cem

en

t

surr

ound

ing

tissu

e un

affe

cted

by

lesi

on

abno

rmal

cal

iber

an

do

r arb

oriz

atio

n st

raig

hten

ing

or th

icke

ning

of C

oope

rs li

gam

ents

(cu

rvili

near

co

nn

ect

ive

tissu

e b

an

ds

prov

idin

g su

ppor

t fo

r th

e br

east

s)

incr

ease

d ec

hoge

nici

ty o

f sur

roun

ding

tis

sue

ret

icul

atio

n i

nclu

des

an

gu

lar

hyp

oe

cho

ic li

nes

disr

uptio

n o

f nor

mal

ana

tom

ic p

lane

s fo

cald

iffus

e sk

in t

hick

enin

g-no

rmal

ski

n is

gt2

mm

in t

hick

ness

exc

ept i

n th

e pe

riar

eola

r ar

ea

and

low

er b

reas

ts

skin

sur

face

is c

onca

ve o

r iII

-def

ined

ap

pear

s pu

lled

in

diso

rder

ed e

cho

pa

tte

m in

volv

ing

pect

oral

mu

scle

sug

gest

ive

of i

nvas

ion

(exc

lude

tech

nica

l ca

uses

suc

h a

s im

prop

erly

pla

ced

foca

l zo

ne

) _

-shy

~ z ~ rJ o N

m

o (lJ

rJ ~ ~

c ~

rJ ampi o c z o r m

x o o z

I

I

C

LA

SS

IFIC

AT

ION

CA

TE

GO

RIE

S amp

TE

RM

S

DE

SC

RIP

TIO

N

Ca

lcif

ica

tio

ns

0 N

one

seen

C

alc

ifica

tion

s a

re p

oo

rly

cha

ract

eri

zed

with

ul

traso

und

but c

an b

e re

cogn

ized

par

ticul

arly

in a

If

pres

ent

(sel

ect a

ll th

at a

pply

) m

ass

sect Macrocalc~ications

Mic

roca

lcifi

catio

ns o

ut o

f m

ass

Mic

roca

lcifi

catio

ns in

mas

s

no c

alci

ficat

ions

see

n

~O5mm

I I I

Sp

ecia

l C

ase

s

0 N

one

Sp

eci

al

case

s ar

e th

ose

with

a u

niq

ue

dia

gn

osi

s (s

elec

t all

that

app

ly)

or

find

ing

~ Mas

s in

or

on s

kin

For

eign

bod

y Ly

mph

nod

es-in

tram

amm

ary

Lym

ph n

odes

-axi

lla

I incl

udin

g se

bace

ous

or e

pide

rmal

incl

usio

n cy

st

kelo

id

etc

in

clud

ing

clip

co

il w

ire c

athe

ter

slee

ve

silic

one

etc

in

bre

ast

incl

udin

g ax

illar

y ta

il

Va

scu

lari

ty

(sel

ect o

ne)

~ Can

not a

sses

s va

scul

arity

N

one

Sam

e as

in n

orm

al t

issu

e D

ecre

ased

In

crea

sed

colo

r flo

w n

ot d

one

or in

adeq

uate

for

inte

rpre

tatio

n no

col

or fl

ow

less

than

in n

orm

al ti

ssue

m

ore

than

in n

orm

al ti

ssue

I

Th

ere

are

lim

ite

d d

ata

to

su

pp

ort

ma

na

ge

me

nt

reco

mm

en

da

tio

ns

tor

so

lid

ma

sse

s b

ase

d o

n u

ltra

so

un

d fi

nd

ing

s a

t th

is t

ime

How

ever

w

hat

wou

ld b

e yo

ur

bes

t ass

essm

ent

an

d m

an

ag

emen

t rec

omm

enda

tion

in e

ach

case

Inco

mp

lete

Asse

ssm

en

t 0

Inco

mpl

ete

addi

tiona

l eva

luat

ion

need

ed b

efor

e fin

al a

sses

smen

t

Fin

al A

sse

ssm

en

t C

ate

go

ry

Neg

ativ

e B

enig

n fin

ding

~ Pro

babl

y be

nign

S

uspi

ciou

s ab

norm

ality

H

ighl

y su

gges

tive

of m

alig

nanc

y ------------------------------shy

no le

sion

foun

d (r

outin

e fo

llow

-up)

no

mal

igna

At f

eatu

res

ex-

cyst

(ro

utin

e fo

llow

-up)

lo

w p

roba

bilit

y o

f can

cer

ex-

fibro

aden

oma

(sho

rt in

terv

al fo

llow

-up)

in

term

edia

te p

roba

bilit

y o

f can

cer

(tis

sue

sam

plin

g)

high

pro

babi

lity

of c

ance

r (t

issu

e sa

mpl

ing)

-

Fo

r e

ach

of

the

ab

ove

ca

teg

ori

es

se

lect

the

te

rm t

ha

t b

est

de

scri

be

s th

e d

om

ina

nt

lesi

on

fe

atu

re

m

Wh

ere

ver

po

ssib

le

de

fin

itio

ns

an

d d

esc

rip

tio

ns

use

d i

n B

I-R

AO

S (

Re

sto

n

VA

) fo

r m

am

mo

gra

ph

y w

ill b

e a

pp

lied

to

ult

raso

un

d

Z o

Ple

ase

ma

rk th

e b

ox

be

sid

e y

ou

r se

lect

ion

m

r (f

JC

opyr

ight

200

1 A

mer

ican

Col

lege

of R

adio

k)gy

Bas

ed o

n F

inal

Rep

ort

of E

xper

t W

orki

ng G

roup

D

evel

oped

Un

de

r Co

ntr

act

282

-97-

0016

-B

etw

een

us

Pub

lic H

eart

h S

ervi

ce O

ffic

e o

n W

omen

s H

ealth

us

De

pa

rtm

en

t o

f Hea

lth

and

Hum

an

o S

ervi

ces

an

d t

he A

me

rica

n C

olle

ge

of R

adiO

logy

z rn

m

IJ

p tgt

Z

o

m

IJ

IJ j

f

_

rr

_

(I)

j _

~

m Q~

-

-

_ ~

0 (

) o

en

en

c--

0 (l

)

shy

~

2

r

c

nC

ll

O-C

sect8sect35~

cr

7

-

r

co

-0

Q)

0 b

~

(1)

-o~

rl

-

~

pl

-

(I)r~m

lt

~

()

223 TOWARD A STANDARDIZED BREAST ULTRASOUND LEXICON

Fig 11 Special cases Skin lesions lymph nodes and foreign bodies are also demonstrable with breast ultrasonogmiddot raphy (A) A typical intramammary lymph node with an echogenic hilum is shown (arrows) (8) Extracapsular spread of silicone gel from a ruptured silicone implant usually manmiddot ifests itself as echogenic noise or a snowstorm pattern (curved arrow)

sing le view of each lesion In prac ti ce descripto rs should be based on multiple views of masses obtained in orthogonal imaging planes in accord with the ACR Standard for the Ultrasound Exa mshyination of the Breast 3 Primary descriptors of masses (shape ori entation margin and echo patshytem ) are generall y listed in order of increasing risk of malignancy top to bottom altho ugh further validation of the risk of malignancy is needed Secshyondary features associated fi ndings or effects on sUlTounding tissue are not listed in any particular order as fLll1her assessment of the risk of mali gnancy for each feature is needed In refening to Tabk 1 it is important to re-emphasize that greatest specific ity is achieved by the evaluation of multiple features of the mass rather than any single attlibute

Certain prob le ms of descrip tion nomenclashyture a nd ca tegori za tion have no t been reso lved

Fig 12 Vascularity The role of vascularity in the characmiddot terization of breast masses is not fully defined Therefore description of vascular features is optional Vascularity asmiddot sessed with color or power Doppler is described as the same increased or decreased compared with adjacent normal breast tissue Power Doppler of a high-grade carcinoma with increased vascularity (arrowheads) is shown

Fig 13 Tubular mass A complex centrally cystic mass tubular in shape with hyperechoic indistinct rim and hetermiddot ogeneous internal echoes is noted Thi s patient is on coumamiddot din and developed a hematoma in the breast Although the features suggest potential need for biopsy the combination of clinical history and imaging findings allows classification of this finding as probably benign with shortmiddotinterval followmiddotup recommended

For example ove rl ap of shape and margin ca teshygories (eg irregular) has bee n discussed in the development of both mammography and ult ra shysound lex icons Cse of the term tu bular may be appropria te as a shape (Fig 13) o r special case (eg dil a ted duct) The benign les io n apocrine metapasia l with its characterist ic m icrolobushyla ted margin bu t otherwise iden ti fi ably benig n microcys tic components (Fig 14) might be a spec ia l case be tter includ ed and desc ri bed for its uni q ue ness than analyzed by indi vi dual desc ri pshyto rs such as irregu lar micro lobu lated and comshyplex that woul d otherwise prompt ti ssue di ag noshySIS

Another dil emma is classifica tion of lesion s by

Fig 14 Apocrine metaplasia (A) Microlobulated masses with microcysts are typical of apocrine metaplasia and may rarely contain calcifications (arrow) (8) Enhancement is sometimes evident Such lesions can be considered benign or probably benign and followed 25

C

LA

SS

IFIC

AT

ION

CA

TE

GO

RIE

S amp

TE

RM

S

DE

SC

RIP

TIO

N

Ca

lcif

ica

tio

ns

0 N

one

seen

C

alc

ifica

tion

s a

re p

oo

rly

cha

ract

eri

zed

with

ul

traso

und

but c

an b

e re

cogn

ized

par

ticul

arly

in a

If

pres

ent

(sel

ect a

ll th

at a

pply

) m

ass

sect Macrocalc~ications

Mic

roca

lcifi

catio

ns o

ut o

f m

ass

Mic

roca

lcifi

catio

ns in

mas

s

no c

alci

ficat

ions

see

n

~O5mm

I I I

Sp

ecia

l C

ase

s

0 N

one

Sp

eci

al

case

s ar

e th

ose

with

a u

niq

ue

dia

gn

osi

s (s

elec

t all

that

app

ly)

or

find

ing

~ Mas

s in

or

on s

kin

For

eign

bod

y Ly

mph

nod

es-in

tram

amm

ary

Lym

ph n

odes

-axi

lla

I incl

udin

g se

bace

ous

or e

pide

rmal

incl

usio

n cy

st

kelo

id

etc

in

clud

ing

clip

co

il w

ire c

athe

ter

slee

ve

silic

one

etc

in

bre

ast

incl

udin

g ax

illar

y ta

il

Va

scu

lari

ty

(sel

ect o

ne)

~ Can

not a

sses

s va

scul

arity

N

one

Sam

e as

in n

orm

al t

issu

e D

ecre

ased

In

crea

sed

colo

r flo

w n

ot d

one

or in

adeq

uate

for

inte

rpre

tatio

n no

col

or fl

ow

less

than

in n

orm

al ti

ssue

m

ore

than

in n

orm

al ti

ssue

I

Th

ere

are

lim

ite

d d

ata

to

su

pp

ort

ma

na

ge

me

nt

reco

mm

en

da

tio

ns

tor

so

lid

ma

sse

s b

ase

d o

n u

ltra

so

un

d fi

nd

ing

s a

t th

is t

ime

How

ever

w

hat

wou

ld b

e yo

ur

bes

t ass

essm

ent

an

d m

an

ag

emen

t rec

omm

enda

tion

in e

ach

case

Inco

mp

lete

Asse

ssm

en

t 0

Inco

mpl

ete

addi

tiona

l eva

luat

ion

need

ed b

efor

e fin

al a

sses

smen

t

Fin

al A

sse

ssm

en

t C

ate

go

ry

Neg

ativ

e B

enig

n fin

ding

~ Pro

babl

y be

nign

S

uspi

ciou

s ab

norm

ality

H

ighl

y su

gges

tive

of m

alig

nanc

y ------------------------------shy

no le

sion

foun

d (r

outin

e fo

llow

-up)

no

mal

igna

At f

eatu

res

ex-

cyst

(ro

utin

e fo

llow

-up)

lo

w p

roba

bilit

y o

f can

cer

ex-

fibro

aden

oma

(sho

rt in

terv

al fo

llow

-up)

in

term

edia

te p

roba

bilit

y o

f can

cer

(tis

sue

sam

plin

g)

high

pro

babi

lity

of c

ance

r (t

issu

e sa

mpl

ing)

-

Fo

r e

ach

of

the

ab

ove

ca

teg

ori

es

se

lect

the

te

rm t

ha

t b

est

de

scri

be

s th

e d

om

ina

nt

lesi

on

fe

atu

re

m

Wh

ere

ver

po

ssib

le

de

fin

itio

ns

an

d d

esc

rip

tio

ns

use

d i

n B

I-R

AO

S (

Re

sto

n

VA

) fo

r m

am

mo

gra

ph

y w

ill b

e a

pp

lied

to

ult

raso

un

d

Z o

Ple

ase

ma

rk th

e b

ox

be

sid

e y

ou

r se

lect

ion

m

r (f

JC

opyr

ight

200

1 A

mer

ican

Col

lege

of R

adio

k)gy

Bas

ed o

n F

inal

Rep

ort

of E

xper

t W

orki

ng G

roup

D

evel

oped

Un

de

r Co

ntr

act

282

-97-

0016

-B

etw

een

us

Pub

lic H

eart

h S

ervi

ce O

ffic

e o

n W

omen

s H

ealth

us

De

pa

rtm

en

t o

f Hea

lth

and

Hum

an

o S

ervi

ces

an

d t

he A

me

rica

n C

olle

ge

of R

adiO

logy

z rn

m

IJ

p tgt

Z

o

m

IJ

IJ j

f

_

rr

_

(I)

j _

~

m Q~

-

-

_ ~

0 (

) o

en

en

c--

0 (l

)

shy

~

2

r

c

nC

ll

O-C

sect8sect35~

cr

7

-

r

co

-0

Q)

0 b

~

(1)

-o~

rl

-

~

pl

-

(I)r~m

lt

~

()

223 TOWARD A STANDARDIZED BREAST ULTRASOUND LEXICON

Fig 11 Special cases Skin lesions lymph nodes and foreign bodies are also demonstrable with breast ultrasonogmiddot raphy (A) A typical intramammary lymph node with an echogenic hilum is shown (arrows) (8) Extracapsular spread of silicone gel from a ruptured silicone implant usually manmiddot ifests itself as echogenic noise or a snowstorm pattern (curved arrow)

sing le view of each lesion In prac ti ce descripto rs should be based on multiple views of masses obtained in orthogonal imaging planes in accord with the ACR Standard for the Ultrasound Exa mshyination of the Breast 3 Primary descriptors of masses (shape ori entation margin and echo patshytem ) are generall y listed in order of increasing risk of malignancy top to bottom altho ugh further validation of the risk of malignancy is needed Secshyondary features associated fi ndings or effects on sUlTounding tissue are not listed in any particular order as fLll1her assessment of the risk of mali gnancy for each feature is needed In refening to Tabk 1 it is important to re-emphasize that greatest specific ity is achieved by the evaluation of multiple features of the mass rather than any single attlibute

Certain prob le ms of descrip tion nomenclashyture a nd ca tegori za tion have no t been reso lved

Fig 12 Vascularity The role of vascularity in the characmiddot terization of breast masses is not fully defined Therefore description of vascular features is optional Vascularity asmiddot sessed with color or power Doppler is described as the same increased or decreased compared with adjacent normal breast tissue Power Doppler of a high-grade carcinoma with increased vascularity (arrowheads) is shown

Fig 13 Tubular mass A complex centrally cystic mass tubular in shape with hyperechoic indistinct rim and hetermiddot ogeneous internal echoes is noted Thi s patient is on coumamiddot din and developed a hematoma in the breast Although the features suggest potential need for biopsy the combination of clinical history and imaging findings allows classification of this finding as probably benign with shortmiddotinterval followmiddotup recommended

For example ove rl ap of shape and margin ca teshygories (eg irregular) has bee n discussed in the development of both mammography and ult ra shysound lex icons Cse of the term tu bular may be appropria te as a shape (Fig 13) o r special case (eg dil a ted duct) The benign les io n apocrine metapasia l with its characterist ic m icrolobushyla ted margin bu t otherwise iden ti fi ably benig n microcys tic components (Fig 14) might be a spec ia l case be tter includ ed and desc ri bed for its uni q ue ness than analyzed by indi vi dual desc ri pshyto rs such as irregu lar micro lobu lated and comshyplex that woul d otherwise prompt ti ssue di ag noshySIS

Another dil emma is classifica tion of lesion s by

Fig 14 Apocrine metaplasia (A) Microlobulated masses with microcysts are typical of apocrine metaplasia and may rarely contain calcifications (arrow) (8) Enhancement is sometimes evident Such lesions can be considered benign or probably benign and followed 25

223 TOWARD A STANDARDIZED BREAST ULTRASOUND LEXICON

Fig 11 Special cases Skin lesions lymph nodes and foreign bodies are also demonstrable with breast ultrasonogmiddot raphy (A) A typical intramammary lymph node with an echogenic hilum is shown (arrows) (8) Extracapsular spread of silicone gel from a ruptured silicone implant usually manmiddot ifests itself as echogenic noise or a snowstorm pattern (curved arrow)

sing le view of each lesion In prac ti ce descripto rs should be based on multiple views of masses obtained in orthogonal imaging planes in accord with the ACR Standard for the Ultrasound Exa mshyination of the Breast 3 Primary descriptors of masses (shape ori entation margin and echo patshytem ) are generall y listed in order of increasing risk of malignancy top to bottom altho ugh further validation of the risk of malignancy is needed Secshyondary features associated fi ndings or effects on sUlTounding tissue are not listed in any particular order as fLll1her assessment of the risk of mali gnancy for each feature is needed In refening to Tabk 1 it is important to re-emphasize that greatest specific ity is achieved by the evaluation of multiple features of the mass rather than any single attlibute

Certain prob le ms of descrip tion nomenclashyture a nd ca tegori za tion have no t been reso lved

Fig 12 Vascularity The role of vascularity in the characmiddot terization of breast masses is not fully defined Therefore description of vascular features is optional Vascularity asmiddot sessed with color or power Doppler is described as the same increased or decreased compared with adjacent normal breast tissue Power Doppler of a high-grade carcinoma with increased vascularity (arrowheads) is shown

Fig 13 Tubular mass A complex centrally cystic mass tubular in shape with hyperechoic indistinct rim and hetermiddot ogeneous internal echoes is noted Thi s patient is on coumamiddot din and developed a hematoma in the breast Although the features suggest potential need for biopsy the combination of clinical history and imaging findings allows classification of this finding as probably benign with shortmiddotinterval followmiddotup recommended

For example ove rl ap of shape and margin ca teshygories (eg irregular) has bee n discussed in the development of both mammography and ult ra shysound lex icons Cse of the term tu bular may be appropria te as a shape (Fig 13) o r special case (eg dil a ted duct) The benign les io n apocrine metapasia l with its characterist ic m icrolobushyla ted margin bu t otherwise iden ti fi ably benig n microcys tic components (Fig 14) might be a spec ia l case be tter includ ed and desc ri bed for its uni q ue ness than analyzed by indi vi dual desc ri pshyto rs such as irregu lar micro lobu lated and comshyplex that woul d otherwise prompt ti ssue di ag noshySIS

Another dil emma is classifica tion of lesion s by

Fig 14 Apocrine metaplasia (A) Microlobulated masses with microcysts are typical of apocrine metaplasia and may rarely contain calcifications (arrow) (8) Enhancement is sometimes evident Such lesions can be considered benign or probably benign and followed 25

l

224

Table 2 BI-RADS Ultrasound Final Assessment Categories

Assessment Categories and Codes Recommendations

O-Incomplete Needs additional imaging evaluation

1-Negative Routine follow-up for age 2-Benign finding Routine follow-up for age 3-Probably benign finding Short-term follow-up

(usually in 6 mo nths) 4-Suspicious abnormality Requires tissue sampling 5-Highly suggestive of Requires tissue sampling

malignancy

their echogenicity that is_ defined relative to fat or fibroglanduar tissue Some masses notably fibroshyadenomas are similar to fat lobules in shape and echogenicity Should they be described as hyposhyechoic using fibroglandular tissue as the basis for comparison just as fat lobules are hypoechoic to fibroglandular tissue or should they be termed isoechoic to fat) Further in need of validation are the categories of thickness of circumscribed marshygins and the characterization of solid masses by their matrix homogeneity 1any fibroadenomas are heterogeneous and some carcinomas are homoshygeneous What is the predictive value of these features alone or in combination with other feashytures

FINAL ASSESSMENT

As with mammography a Br-RADS final asshysessment and recommendation should be specified (Table 2) When breast sonography is pelformed as an adjunct to mammography one final as sessment and management recommendation should be specshyified which reflects combined mammographic and sonographic findings Final assessment and man-

MENDELSON BERG AND MERRITI TOW

agement should be predicated on the basis of the II

most suspicious feature(s) present preo Radl

I SUMMARY

solie

The approach outlined above for describing and and

reporting sonographic features of breast masses I

represents only the initial step in the development repc m

of a comprehensive system to enhance the accurate I

identification reporting and analysis of sonoshygraphic abnormalities of the breast Future revishysions with validation of interobserver consistency anal

in application of these descriptors across multiple toll

I centers with feedback from potential users in the Cha breast imaging community will undoubtedly exshy M O

pand the utility of this eff0l1 I bre

AJFACKNOWLEDGMENTS I

Participating in the development of the standardshy son

ized breast lexicon were the members of the Expert Working Group to Plan and Develop Protocols for WOI

Optimization and Clinical Testing of Breast Ultrashydial

19S sound Steering Committee-Christopher R B Menitt MD Chairman Barry B Goldberg MD con

Ellen B 1endel son MD Lexicon Subcommitshy 191

tee-Ellen B 1endelson MD chair Janet K Baum MD Christopher R B Merritt MD Eva

Inc

Rubin MD ExpeI1 Working Group-Janet K Baum MD Lawrence W Bassett MD Paul L Carson_ PhD David Cosgrove MD A Pat Romshyilly 10 Barbara S Hertzberg 10 Valerie P Jackson 10 Robert R Kuske Jr MD Helmut Madjar MD PhD Jonathan Ophir PhD Steve H Parker MD Catherine W Piccoli MD Eva Rushybin MD Gordon F Schwartz MD Edward A Sickles MD and Jonathan Sunshine PhD

REFERENCES

1 Rosenberg AL Schwanz GF Feig SA et al Clinically

occult breast les ions Localization and signincanclt Radiology

162 167-170 1987

2 Bassett LW Liu TH Giuliano AE et al The prevalence

of carcinoma in palpable vs impalpable mammog raphicaJly

detected les ions AJR Am J Roentgenol 157 21-24 1991

3 ACR Standard for the Pelformance of Breast Ultrasou nd

Examination ACR Standards Resto n V A American Coll ege

of Radiology 2000 pp 389-392

4 Merritt CRB The Breast Nodule in Ultrasound A Pracshy

tical Approach to Clinica l Problems (ch 52) in Bluth EI Ar2er

P Benson C et al (eds) A Prac ti ca l Approach to Clinical

Problems New York Thieme 2000

5 Stavros AT Thickman D Rapp CL et a1 Solid breast

nodules Use of sonog raphy to di stingui sh between beni g n and

malignant lesions Radiology 196 123-134 1995

6 Taylor KJW HOI Breast Sc reening Researc h (ATL) Can

complementary US reduce the number of biopsies of beni gn

breast massesry Radiology 189(P) 179 1993 (abstr suppl )

7 High defi niti on imag ing The ro le o f ultrasound in the

di agnosis of breast cancer Bothell WA Advanced Technology

Labo ratorie 1 99 ~

8 Zonderand HM Coerkamp EG Hem1ans J et al Diagshy

nosis of breast cancer Contribution of US as an adjuJlc t to

mammography Radiol ogy 21341 3-422 1999

9 Sickl es EA Filly RA Callen PW Breast cancer detecshy

ti o n wi th so nograph y a nd mammography Compariso n using

sta te-o f-the-a rt equipme nt AJR Am J Roenlgenol 140843shy

84 5 1983

10 Skaane P Sauer T Ultrasonograph y of malignant breast

neoplasms Anal ys is of cuc inomas missed as lumor Acta

Radiol 40376-382 1999

225 TOWARD A STANDARDI ZED BREAST ULTRASOUND LEXICON

II Berg W A Gilbrea th PL Whole breast ul trasound in preoperative evaluati on for multicentric and multi foca l cancer Radiology 2 1459-66 2000

12 Baker lA Kornguth Pl Soo MS et al Sonography of solid breast les ions Observer variability of les ion description and assessment AJR Am 1 Roentgenol 172 1621- 1625 1999

13 American Co llege of Radiology (ACR) Breast imag ing reponing and data system (BI-RADSlgtt) (ed 3) Res ton VA Ameri can College of Radiology 1998

14 Morris EA lliustrated breast MR lex icon Semin Roentshygenol 36XX 2001

15 Cole-Beuglet C Soriano RZ Kurtz B et al Ultrasound analysis of 104 primary breast carcinomas c1ass itied according to hi stopathologic type Radio logy 147 1 91-196 1983

16 Mendelson EB The Breast in Wilson S Rumack C Charboneau JW (eds ) Diagnost ic Ultrasound (ed 2) St Louis Mosby 1998 pp 75 1-789

17 Hilton SvW Leopold GR Ol son LK et al Real-time breast sonography Applica ti on in 300 consecuti ve patients AJR Am J Roentgenol 147479-486 1986

18 Jokich PM Monticciolo DL dler YT Breast ultrashysonography Radi ol Cl in North Am 30993-1009 1992

19 Kolb TM Lichy J Newhouse JH Occu lt cancer in women with dense breasts Detection with sc reening USshydi agnostic yield and tumor charac teri stics Rad iology 207 191 shy1991998

20 Venta LA Kim JP Pelloski CE et al Management of complex breast cysts AJR Am J Roentgenol 173 133 1-1 336 1999

21 Buchberger W DeKoek koek-Do ll P Springer P ct al Incidental findings on sonog raphy of the breast Clinical sigshy

ni fi cance and diagnostic worku p AJ R Am J Roentgenol 173 92 1-927 1999

22 Jackson VP Management of so lid breast nodul es What is the role of sonography~ Radiology 196 14- 15 1995 (editoshyrial )

23 Sick les EA Periodic mammographic fo ll ow-u p of prob shyab ly benign les ions Results in 3184 consecuti ve cases Radishyology 179463-468 1991

24 Sickles EA )onpalpable c ircumscribed nonca lcified so lid breast masses Like lihood of ma lignancy based on les ion size and age of patient Radiology J92439-442 1994

25 Warner lK Kum Jr 0 Berg WA Apocrine metaplas ia Mammographic and sonographic appearances AJ R m J Roentgenol 170 1375-1379 1998

26 Rahbar G Sie Ac Hansen Gc et a1 Benign versus malignant sol id breas t masses US different iation Radiology 2 13889-894 1999

27 Hanis KJv Ganon MA Shestak KC et al Si licone implant rupture Detect ion wi th US Radiology 18776 1-768 1993

28 Cosgrove DO Kedar RP Bamber Jc et al Breast di seases Color doppler US in differenti al diagnosis Radiology 18999-104 1993

29 VlcNic holas MM Mercer PM Miller JC et al Color Doppler sonography in the evaluation of palpable breast masses AJR m 1 Roentgenol 161 765-77 1 1993

30 Ra7a S Ballm JK Solid breast les ions Evaluation with power Doppler tS Radiology 203 164-168 1997

31 Skaane P Engedal K Anal ysis of sonographic features in the differentiation of fibroadenoma and invas ive ducta l carcino ma AJR m J Roentgenol 170 109- 1 14 1998

225 TOWARD A STANDARDI ZED BREAST ULTRASOUND LEXICON

II Berg W A Gilbrea th PL Whole breast ul trasound in preoperative evaluati on for multicentric and multi foca l cancer Radiology 2 1459-66 2000

12 Baker lA Kornguth Pl Soo MS et al Sonography of solid breast les ions Observer variability of les ion description and assessment AJR Am 1 Roentgenol 172 1621- 1625 1999

13 American Co llege of Radiology (ACR) Breast imag ing reponing and data system (BI-RADSlgtt) (ed 3) Res ton VA Ameri can College of Radiology 1998

14 Morris EA lliustrated breast MR lex icon Semin Roentshygenol 36XX 2001

15 Cole-Beuglet C Soriano RZ Kurtz B et al Ultrasound analysis of 104 primary breast carcinomas c1ass itied according to hi stopathologic type Radio logy 147 1 91-196 1983

16 Mendelson EB The Breast in Wilson S Rumack C Charboneau JW (eds ) Diagnost ic Ultrasound (ed 2) St Louis Mosby 1998 pp 75 1-789

17 Hilton SvW Leopold GR Ol son LK et al Real-time breast sonography Applica ti on in 300 consecuti ve patients AJR Am J Roentgenol 147479-486 1986

18 Jokich PM Monticciolo DL dler YT Breast ultrashysonography Radi ol Cl in North Am 30993-1009 1992

19 Kolb TM Lichy J Newhouse JH Occu lt cancer in women with dense breasts Detection with sc reening USshydi agnostic yield and tumor charac teri stics Rad iology 207 191 shy1991998

20 Venta LA Kim JP Pelloski CE et al Management of complex breast cysts AJR Am J Roentgenol 173 133 1-1 336 1999

21 Buchberger W DeKoek koek-Do ll P Springer P ct al Incidental findings on sonog raphy of the breast Clinical sigshy

ni fi cance and diagnostic worku p AJ R Am J Roentgenol 173 92 1-927 1999

22 Jackson VP Management of so lid breast nodul es What is the role of sonography~ Radiology 196 14- 15 1995 (editoshyrial )

23 Sick les EA Periodic mammographic fo ll ow-u p of prob shyab ly benign les ions Results in 3184 consecuti ve cases Radishyology 179463-468 1991

24 Sickles EA )onpalpable c ircumscribed nonca lcified so lid breast masses Like lihood of ma lignancy based on les ion size and age of patient Radiology J92439-442 1994

25 Warner lK Kum Jr 0 Berg WA Apocrine metaplas ia Mammographic and sonographic appearances AJ R m J Roentgenol 170 1375-1379 1998

26 Rahbar G Sie Ac Hansen Gc et a1 Benign versus malignant sol id breas t masses US different iation Radiology 2 13889-894 1999

27 Hanis KJv Ganon MA Shestak KC et al Si licone implant rupture Detect ion wi th US Radiology 18776 1-768 1993

28 Cosgrove DO Kedar RP Bamber Jc et al Breast di seases Color doppler US in differenti al diagnosis Radiology 18999-104 1993

29 VlcNic holas MM Mercer PM Miller JC et al Color Doppler sonography in the evaluation of palpable breast masses AJR m 1 Roentgenol 161 765-77 1 1993

30 Ra7a S Ballm JK Solid breast les ions Evaluation with power Doppler tS Radiology 203 164-168 1997

31 Skaane P Engedal K Anal ysis of sonographic features in the differentiation of fibroadenoma and invas ive ducta l carcino ma AJR m J Roentgenol 170 109- 1 14 1998