26.6.jum.online.

q 5/16/07 2:15 PM Page 807

Article

Comparison of Ultrasound Elastography,

Mammography, and Sonography in the
Diagnosis of Solid Breast Lesions

Hui Zhi, MD, Bing Ou, MD, Bao-Ming Luo, MD,

Xia Feng, MD, Yan-Ling Wen, MD, Hai-Yun Yang, MD

Objective. The purpose of this study was to evaluate the value of ultrasound elastography (UE) in dif-
ferentiating benign versus malignant lesions in the breast and compare it with conventional sonogra-
phy and mammography. Methods. From September 2004 to May 2005, 296 solid lesions from 232
consecutive patients were diagnosed as benign or malignant by mammography and sonography and
further analyzed with UE. The diagnostic results were compared with histopathologic findings. The
sensitivity, specificity, accuracy, positive and negative predictive values, and false-positive and -negative
rates were calculated for each modality and the combination of UE and sonography. Results. Of 296
lesions, 87 were histologically malignant, and 209 were benign. Ultrasound elastography was the most
specific (95.7%) and had the lowest false-positive rate (4.3%) of the 3 modalities. The accuracy
(88.2%) and positive predictive value (87.1%) of UE were higher than those of sonography (72.6%
and 52.5%, respectively). The sensitivity values, negative predictive values, and false negative rates of
the 3 modalities had no differences. A combination of UE and sonography had the best sensitivity
(89.7%) and accuracy (93.9%) and the lowest false-negative rate (9.2%). The specificity (95.7%) and
positive predictive value (89.7%) of the combination were better, and the false-positive rate (4.3%) of
the combination was lower than those of mammography and sonography. Conclusions. In a clinical
trial with Chinese women, UE was superior to sonography and equal or superior to mammography in
differentiating benign and malignant lesions in the breast. A combination of UE and sonography had
the best results in detecting cancer and potentially could reduce unnecessary biopsy. Ultrasound elas-
tography is a promising technique for evaluating breast lesions. Key words: breast carcinoma; mam-
mography; sonography; ultrasound elastography.

B
Abbreviations reast cancer is the most common malignancy in
ROI, region of interest; UE, ultrasound elastography
women and the second most common cause of
cancer-related mortality.1 The most current esti-
mates from the International Agency for Research
on Cancer for the global disease burden of breast cancer
are for 2002, and in that year, the agency estimated that
Received December 4, 2006, from the Department
of Ultrasound, Second Affiliated Hospital, Sun Yat-
there were approximately 1.15 million newly diagnosed
Sen University, Guangzhou China. Revision request- cases and approximately 411,000 deaths.2 On the basis of
ed January 17, 2007. Revised manuscript accepted current estimates of an average annual increase in inci-
for publication February 5, 2007.
Address correspondence to Bao-Ming Luo, MD, dence ranging from 0.5% to 3% per year, the number of
Department of Ultrasound, Second Affiliated new cases projected to be diagnosed in 2010 is 1.4 to 1.5
Hospital of Sun Yat-Sen University, 107 Yanjiangxi
Rd, 510120 Guangzhou, China.
million.2 Recent research on service screening programs
E-mail: baomingluo2005@126.com suggests that participation in modern, organized service

© 2007 by the American Institute of Ultrasound in Medicine • J Ultrasound Med 2007; 26:807–815 • 0278-4297/07/$3.50
26.6.jum.online.q 5/16/07 2:15 PM Page 808

Ultrasound Elastography in Solid Breast Lesions

screening may well reduce the risk of dying of The strain map of ultrasound elastography
breast cancer by 40% or more.3 That is, early and (UE) is superimposed on a conventional B-mode
sensitive diagnosis represents a better prognosis. sonogram. The goal of this study was to evaluate
Noninvasive diagnosis of breast cancer remains whether the new method of UE could improve
a major clinical problem. Mammography and the differentiation and characterization of
sonography are currently the most sensitive benign and malignant breast lesions compared
modalities for detecting breast cancer. Most with conventional sonography and mammogra-
Chinese women have relatively small, dense phy and whether the combination of UE and
breasts,4 which is one of the various factors lead- sonography could improve the accuracy of diag-
ing to false-negative findings on mammography.5 nosis of breast lesions significantly.
Practically, sonography is chosen as the primary
workup tool in the clinic in China. However, the Materials and Methods
sonographic features for benign and malignant
lesions have been shown to override each other Patients
substantially.6,7 These limitations of mammogra- From September 2004 to May 2005, 232 patients
phy and sonography and the great desire not to with 296 solid lesions in the breast were includ-
miss a malignant lesion in the early stage of dis- ed in the study. The mean age of the patients
ease lead to aggressive biopsy, but the biopsy was 42 years, with a range of 17 to 87 years. Final
rate for cancer is only 10% to 30%.8,9 This means diagnosis was confirmed by histopathologic
that 70% to 90% of breast biopsies are per- examination. Informed consent for diagnostic
formed for benign diseases, which induce procedures was obtained from each patient.
unnecessary patient discomfort and anxiety in The study was approved by the Ethics
addition to increasing costs to the patient. Committee of the hospital. All patients were
Clearly, there is a great need for development of examined with mammography and sonogra-
additional reliable methods to complement the phy, and those with solid lesions were further
existing diagnostic procedures to avoid unnec- analyzed with UE. All the examinations were
essary biopsy. performed before any surgery, biopsy, or fine-
In the clinical examination, palpation is the needle aspirations. Only the patients with all of
standard screening procedure for the detection these examination findings available were
of breast, thyroid, prostate, and liver abnormali- included in the study.
ties. That is done on the basis that cancer tissue
is harder than adjacent normal tissue. However, Study Design
palpation is not very accurate because of its poor Mammography was performed with a
sensitivity as well as its limited accuracy in terms Senographe 600T mammography system (GE
of different locations of lesions. A better under- Healthcare, Milwaukee, WI). Standard medio-
standing of soft tissue elasticity, another charac- lateral oblique and craniocaudal projections
teristic for delineating the properties of a lesion were obtained. Additional projections, such as
in addition to morphologic characteristics that coned-down compression and magnification
can be detected with sonography and other views, were considered when further help in
conventional radiologic modalities, and a sys- analysis was expected. Mammograms were
tem for accurately and efficiently predicting it reviewed for masses, calcifications, architec-
are therefore needed. In the early 1990s, a tech- tural distortion, and tubular ductal opacity. The
nique called elastography was described by shape, margin, and density of masses were
Ophir et al.10 With this technique, the tissue is noted. Two different radiologists with at least 5
compressed, and the tissue strain resulting from years of experience in breast mammography
this compression is imaged. Since its invention, read the mammograms and made the diagnosis
this concept has been proposed for elasticity of the lesions together.
imaging of a wide range of different applica- In each patient, bilateral whole-breast sonog-
tions, including prostate,11 breast,12–15 thyroid,16 raphy was performed in the transverse and lon-
and intravascular ultrasound.17 gitudinal planes with a EUB-8500 ultrasound

808 J Ultrasound Med 2007; 26:807–815

26.6.jum.online.q 5/16/07 2:15 PM Page 809

Zhi et al

scanner (Hitachi Medical Corporation, Tokyo, Data Analysis

Japan) equipped with a 7.5- to 13.0-MHz linear Interpretation of the various diagnostic proce-
array transducer. Sonographic features were dures was compared with the histologic findings
prospectively recorded in a computer database with regard to sensitivity, specificity, accuracy,
and included anteroposterior-width ratio, positive and negative predictive values, and
shape, margin, internal echogenicity, internal false-positive and -negative rates. Sensitivity,
echo texture, and posterior acoustic phenome- specificity, negative and positive predictive val-
na. With these features, 2 different sonogra- ues, and accuracy were evaluated as follows:
phers with at least 5 years of experience in sensitivity = patients with suspected breast can-
breast sonography diagnosed the lesions as cer/patients with histologically confirmed breast
benign or malignant together. cancer; specificity = patients with suspected
Ultrasound elastography was performed at benign disease/patients with histologically con-
the same time as the sonographic examination. firmed benign disease; positive predictive value =
Because the ultrasound scanner was equipped patients with histologically confirmed breast can-
with an elastography unit, the elasticity of a cer/patients with suspected breast cancer; nega-
lesion could be measured by a different color. tive predictive value = patients with histologically
On the B-mode image, we displayed the target confirmed benign disease/patients with suspect-
lesion. Then we moved the region of interest ed benign disease; and accuracy = patients with
(ROI) around the lesion, making sure that the true-positive and -negative detected disease/all
target tissue occupied no more than one third patients with lesions (benign and malignant).
of the total area of the ROI. The probe was A result was classified as false-negative when a
moved inferior and superior to obtain the elas- diagnostic method classified histologically con-
ticity images. Importantly, to obtain images firmed cancer as benign. A result was classified
that were appropriate for analysis, we applied as false-positive when a diagnostic method clas-
the probe with only light pressure, with the sified a histologically confirmed benign lesion as
pressure indicator bar displayed as only 2 to 3. cancer. We obtained the false-positive and -neg-
The target lesion was scored as 1 to 5, using the ative rates and compared the performance of all
scoring system proposed by Itoh et al.14 If a diagnostic methods individually and the combi-
lesion was scored as 1 to 3, it would be benign; nation of sonography and UE for all patients.
if a lesion was scored as 4 to 5, it would be
malignant. A score of 1 indicated even strain Statistical Analysis
for the entire hypoechoic lesion (ie, the entire Statistical analysis was performed for all vari-
lesion was evenly shaded in green). A score of ables with the McNemar test and the Pearson χ2
2 indicated strain in most of the hypoechoic test. P < .05 was considered statistically signifi-
lesion with some areas of no strain (ie, the cant. All the statistical analyses in this study were
hypoechoic lesion had a mosaic pattern of carried out with SPSS for Windows software
green and blue). A score of 3 indicated strain at (SPSS Inc, Chicago, IL).
the periphery of the hypoechoic lesion, with
sparing of the center of the lesion (ie, the Results
peripheral part of lesion was green, and the
central part was blue). A score of 4 indicated no All patients underwent breast surgery, and all
strain in the entire hypoechoic lesion (ie, the abnormal lesions identified by mammography,
entire lesion was blue, but its surrounding area sonography, and UE were surgically removed. A
was not included). A score of 5 indicated no total of 296 breast lesions were examined histo-
strain in the entire hypoechoic lesion or in the logically. Histologic analysis showed that 209
surrounding area (ie, both the entire hypoe- (70.6%) of 296 lesions were benign and 87
choic lesion and its surrounding area were (29.4%) of 296 lesions were malignant. The histo-
blue). With the scoring system, 2 different logic diagnoses are summarized in Table 1.
sonographers gave the diagnosis of the lesions Lesion sizes ranged from 3.1 to 100.6 mm in
as benign or malignant together. maximum diameter (mean, 16.5 mm; median,

J Ultrasound Med 2007; 26:807–815 809

26.6.jum.online.q 5/16/07 2:15 PM Page 810

Ultrasound Elastography in Solid Breast Lesions

Table 1. Histologic Diagnoses of Benign and Malignant for mammography was significantly higher
Breast Lesions in 232 Patients than that of sonography (P < .05; Table 2).
Histopathologic Diagnosis n Combining UE and sonographic diagnostic
methods yielded the best results for detection of
Benign lesions (n = 209)
Fibroadenoma 125 cancer (Table 2). The highest results for accura-
Fibrocystic mastopathy 57 cy and sensitivity were seen for a combination
Papilloma 12 of the 2 methods (P < .05; Table 2). The specifici-
Chronic inflammation 7
Mucous adenoma 1
ty and positive predictive value of the combina-
Hyperplasia 6 tion were significantly higher than those of
Hematoma 1 mammography and sonography (P < .05; Table
Malignant lesions (n = 87) 2). The negative predictive value for the combi-
Invasive ductal carcinoma 66
Ductal carcinoma in situ 10 nation was significantly higher than that of
Cystosarcoma phyllodes 6 sonography (P < .05; Table 2).
Mucinous carcinoma 1 Mammographic findings were false-negative
Paget disease 1
in 24 of 87 cancers; sonographic findings were
Medullary carcinoma 1
Invasive lobular carcinoma 1 false-negative in 25 of 87 cancers; and 26 of 87
Papillocarcinoma 1 cancers were missed by UE. With regard to the
false-negative rate, no significant difference
between the 3 modalities was found. Most false-
negative findings on UE were found in early
13.3 mm; SD, 11.2 mm); in particular, the benign stages of invasive ductal carcinoma, which were
lesions were 3.1 to 100.6 mm (mean, 13.9 mm; all in stages 1 and 2, and noninvasive carcinoma
median, 11.7 mm; SD, 9.6 mm), and the malig- (Table 3). Six of the false-negative invasive ductal
nant lesions were 5.7 to 58.6 mm (mean, 23.1 carcinomas had somewhat large central necro-
mm; median, 20.2 mm; SD, 12.3 mm). sis, and 5 of the 6 cystosarcoma phyllodes had
With regard to sensitivity and negative predic- false-negative findings on UE (Table 3). Of 26
tive value, no significant difference between the cancers missed by UE, 18 were detected by
3 modalities was found (Table 2). The specifici- sonography. When the diagnoses from UE and
ty of UE was significantly higher than those of sonography were combined, the false-negative
mammography and sonography (P < .05; Table rate was 8 of 87 cancers, which was significantly
2). The specificity of mammography was signif- lower than that of any 1 of the 3 diagnostic meth-
icantly higher than that of sonography (P < .05; ods. The pathologic findings of the 8 missed can-
Table 2). The accuracy rates of UE and mam- cers included 5 invasive ductal carcinomas, 1
mography were significantly higher than that of papillocarcinoma, 1 ductal carcinoma in situ,
sonography (P < .05; Table 2). The positive pre- and 1 Paget disease. Four of the 8 were detected
dictive value for UE was significantly higher by mammography, and 4 were missed by all 3
than those of mammography and sonography modalities, including 2 invasive ductal carcino-
(P < .05; Table 2). The positive predictive value mas, 1 papillocarcinoma, and 1 Paget disease.

Table 2. Comparison of Sensitivity, Specificity, Accuracy, and Positive and Negative Predictive Values for
Mammography, B-Mode Sonography, and UE in the Differentiation of Benign From Malignant Breast Cancer
Imaging Modality Sensitivity, % (n) Specificity, % (n) Accuracy, % (n) PPV, % (n) NPV, % (n)
Mammography 72.4 (63/87) 87.1 (182/209)* 82.7 (245/296)* 70.0 (63/90)* 88.3 (182/206)
Sonography (B-mode) 71.2 (62/87) 73.2 (153/209) 72.6 (215/296) 52.5 (62/118) 86.0 (153/178)
UE 70.1 (61/87) 95.7 (200/209)*† 88.2 (261/296)* 87.1 (61/70)*† 88.5 (200/226)
Sonography (B-mode) + UE 89.7 (78/87)*†‡ 95.7 (200/209)*† 93.9 (278/296)*†‡ 89.7 (78/87)*† 95.7 (200/209)*
NPV indicates negative predictive value; and PPV, positive predictive value.
*P < .05 versus sonography.
†P < .05 versus mammography.
‡P < .05 versus UE.

810 J Ultrasound Med 2007; 26:807–815

26.6.jum.online.q 5/16/07 2:15 PM Page 811

Zhi et al

The false-positive rates of mammography, from breast cancer and the extent of treatment
sonography, and UE were 12.9% (27/209), 26.8% required is early detection through screening.
(56/209), and 4.3% (9/209), respectively. The Ultrasound elastography is a new screening
false-positive rate of UE was significantly lower modality in addition to sonography for detecting
than that of mammography and sonography. and identifying lesions in the breast. It can pro-
The histologic types of the false-positive find- vide the investigator with another characteristic,
ings on UE are listed in Table 3. Among the false- stiffness, of the lesion. Through lightly compress-
positive findings, 2 of fibroadenoma were with ing of the target lesion, UE can noninvasively
calcifications, and 1 of fibrocystic mastopathy determine strain and elasticity distributions
was with dotted, organized remote hemorrhage. inside objects scanned and map the elasticity of
When the diagnoses from UE and sonography the lesion by using a standardized color scale,
were combined, the false-positive was 9 of 209 with blue indicating regions with low elasticity
benign lesions, which was the same as for UE (harder tissue areas) and red indicating regions
alone. with high elasticity (soft tissue).
An illustration of a benign lesion and a malig-
nant lesion detected by UE that had almost the
same benign morphologic characteristics on
sonography is shown in Figure 1. An illustration
of a benign lesion and a malignant lesion detect- Figure 1. Two lesions with almost the same benign morphologic
ed by UE that had almost the same malignant characteristics. A, Fibroadenoma in a 35-year-old woman. Right,
Sonographic image. A hypoechoic mass with a regular shape was
morphologic characteristics on sonography is diagnosed as a benign lesion. Left, Ultrasound elastographic
shown in Figure 2. An illustration of a benign image. The lesion had a mosaic pattern of green and blue, was
lesion detected by UE that was falsely diagnosed scored 1, and was diagnosed as a benign lesion. B, Invasive duc-
tal carcinoma in a 45-year-old woman. Right, Sonographic image.
as cancer by mammography is shown in Figure
A hypoechoic mass with a regular shape was diagnosed as a
3. An illustration of cancer detected by UE that benign lesion. Left, Ultrasound elastographic image. The entire
was missed by mammography is shown in lesion was blue, was scored 4, and was diagnosed as a malignant
Figure 4. lesion.

A
Discussion

Breast cancer is the most common malignancy

among women worldwide.1 In the absence of a
known preventable cause of breast cancer, the
single most important factor in reducing death

Table 3. Analysis of False-Negative and -Positive

Diagnoses With UE
Histopathologic Diagnosis n B
False-negative (n = 26)
Invasive ductal carcinoma 12
Ductal carcinoma in situ 6
Cystosarcoma phyllodes 5
Mucinous carcinoma 1
Papillocarcinoma 1
Paget disease 1
False-positive (n = 9)
Fibroadenoma 3
Fibrocystic mastopathy 4
Papilloma 1
Hyperplasia 1

J Ultrasound Med 2007; 26:807–815 811

26.6.jum.online.q 5/16/07 2:15 PM Page 812

Ultrasound Elastography in Solid Breast Lesions

To our knowledge, there are few reports about invasive carcinoma having the lowest elasticity,
UE for diagnosis of breast lesions, especially the followed by noninvasive carcinoma, fibrous tis-
color-scaled UE that was evaluated in our study. sue in the breast, normal glandular breast tis-
Itoh et al14 first used UE to detect breast lesions sue, and breast fat tissue in that order. In our
and proposed the 5-point scoring system. They study, we found that the specificity, accuracy,
had higher sensitivity of UE than that of con- and positive predict value for the combination
ventional sonography. Thomas et al15 evaluated of sonography and UE were higher than those
this new modality in 108 patients and found of sonography, which is the first choice of
that specificity was improved from 78% for
conventional sonography to 91.5% for UE.
Figure 3. Fibroadenoma in a 39-year-old woman. A,
Our results suggest that UE is a highly sensi-
Mammogram showing a mass suspected to be a malignant
tive, accurate means of identifying and assess- lesion. B, Right, Sonographic image. A hypoechoic mass with an
ing solid lesions of the breast. Our results are in irregular shape was suspected to be a malignant lesion. Left,
agreement with those in a 1998 study by Ultrasound elastographic image. The lesion had a mosaic pat-
tern of green and blue, was scored 2, and was diagnosed as a
Krouskop et al,18 which showed that various benign lesion.
breast tissues had differing elastic stiffness,
A

Figure 2. Two lesions with almost the same malignant mor-

phologic characteristics. A, Papilloma in a 49-year-old woman.
Right, Sonographic image. A hypoechoic mass with an irregular
shape was diagnosed as a malignant lesion. Left, Ultrasound
elastographic image. The lesion had a mosaic pattern of green
and blue, was scored 3, and was diagnosed as a benign lesion.
B, Invasive ductal carcinoma in a 42-year-old woman. Right,
Sonographic image. A hypoechoic mass with an irregular shape
was diagnosed as a malignant lesion. Left, Ultrasound elasto-
graphic image. The entire lesion was blue, was scored 4, and
was diagnosed as a malignant lesion.

B B

812 J Ultrasound Med 2007; 26:807–815

26.6.jum.online.q 5/16/07 2:15 PM Page 813

Zhi et al

modality for screening the breasts of Chinese women with dense breasts and in women
woman. Most Chinese women have relatively younger than 50 years. One of the various fac-
small, dense breasts,4 and the median age of tors leading to false-negative findings on mam-
these patients with breast cancer is 47 years, mography is the effect of breast density.5
which is 8 years younger than in Western Furthermore, breast density on mammography
women,19 which limits the use of mammogra- is also associated with an increased risk of
phy. Bird et al20 stated that breast cancer was breast cancer.21 Meanwhile, dense glandular
less likely to be detected by mammography in tissue usually has a hyperechoic appearance
on sonography, and mostly breast cancers are
Figure 4. Invasive ductal carcinoma in a 55-year-old woman. hypoechoic; carcinomas in this setting are eas-
A, Mammogram showing several masses suspected to be ily detected on sonography. In our study, we
benign lesions. B, Right, Sonographic image. A hypoechoic mass found that UE was better than sonography for
with a round shape was suspected to be a benign lesion. Left,
Ultrasound elastographic image. Both the entire hypoechoic
detecting breast cancer in Chinese patients.
lesion and its surrounding area were blue. The blue area on UE Compared with mammography, the specificity
was larger than the lesion area on sonography. It was scored 5 of UE was higher, but there was no significant
and was diagnosed as a malignant lesion.
difference between the sensitivity, accuracy,
and positive and negative predictive values of
A
UE and mammography. However, mammogra-
phy uses ionizing radiation, which itself is a
potential carcinogen,22 and this limits the age
and frequency with which it can be used.23
Conversely, there is no such risk to patients
when using UE. In addition, UE and sonogra-
phy are much less expensive than mammogra-
phy in China.
There is an overlap of the elasticity between
benign and malignant lesions in the breast,13
which limits the use of UE. In our study, 26 of
87 cancers were missed by UE. Most false-neg-
ative findings on UE were found in early stages
of invasive ductal carcinoma, which were all in
stages 1 and 2, and in noninvasive carcinoma,
and some invasive soft tissue carcinomas such
as cystosarcoma phyllodes6 had large central
necrosis (the lesions with large central necrosis
all had false-negative findings). Consequently,
large-scale necrosis may impair the diagnostic
assessment in UE.15 Nine of 209 benign lesions
were misdiagnosed by UE. That rate was much
B lower than those for mammography and
sonography, which would decrease unneces-
sary biopsies considerably. Among the false-
positive diagnoses, 3 had calcifications, which
might affect the diagnosis on UE, and 1 had a
dotted remote hemorrhage that had already
been organized, which may have increased the
hardness of the lesion. Therefore, when using
UE, one should pay attention to all the factors
that would affect the stiffness of lesions and
cause misleading results.

J Ultrasound Med 2007; 26:807–815 813

26.6.jum.online.q 5/16/07 2:15 PM Page 814

Ultrasound Elastography in Solid Breast Lesions

Ultrasound elastography is superimposed on 9 could be diagnosed by mammography, includ-

conventional sonography. There have been ing 4 invasive ductal carcinomas, 4 ductal carci-
marked advances in the quality of ultrasound nomas in situ, and 1 mucinous carcinoma.
imaging over the past 2 decades, and the ability When using UE to screen a lesion, there are
to differentiate benign and malignant features of many things to pay attention to. The first is that
solid masses with the use of ultrasound has the area of the target lesion occupied in the ROI
improved. This was achieved because of techno- should be less than one third. The echo signals
logical improvements. When we combined the 2 acquired with the ultrasound scanner are cap-
modalities together, we got the best results for tured on an external computer and used to cal-
detection of breast cancer. The sensitivity, speci- culate the tissue strain with the combined
ficity, accuracy, and positive predictive value autocorrelation method.27 This is a qualitative
improved to 89.7%, 95.7%, 93.9%, and 89.7%, elasticity measurement that is relative to the
respectively, which were much higher than those average strain inside the ROI. To compare the tar-
of mammography and sonography. In addition, get lesion elasticity with that of normal breast tis-
the negative predictive value improved to 95.7%, sue, both tissue types should be present in the
which was higher than that of sonography, and ROI.28 The second thing is that light pressure on
the false-negative rate dropped to 8 of 87 cancers, the target lesion should be applied with the
which was much lower than those of mammog- probe by hand. There is a pressure indicator on
raphy (24/87) and sonography (25/87). This the machine, and the pressure scale should be
combination includes detection of 2 features of a between 2 and 3. If the pressure scale is greater
lesion, morphologic characteristics and hard- than 3, it means that the pressure is too high,
ness, which reflect the properties of the lesion. which causes nonlinear properties of tissue elas-
Shape, margin, internal echogenicity, echo tex- ticity and leads to misdiagnosis.
ture distribution, posterior echo, and a bilateral In conclusion, UE is superior to conventional
reflection sign have been shown to be valuable in sonography and is superior or equal to mam-
the differential diagnosis of benign and malig- mography in differentiating benign and malig-
nant breast tumors.24 Conversely, malignant tis- nant lesions in the breast. By combining UE and
sue usually is harder than normal breast tissue. sonography, the detection accuracy can be
Therefore, the 2 elements, morphologic charac- improved greatly, and the combination poten-
teristics and hardness, can reflect whether the tially could reduce unnecessary biopsy. This
lesion is benign or malignant. This combination, combination was the optimal screening modali-
however, did not affect the false-positive rate, ty in our study. Ultrasound elastography is a
which was the same as that for UE alone. promising technique for differentiation of breast
Although there are many limits for the use of lesions.
mammography in Chinese woman, every modal-
ity has its own advantages. Mammography is use- References
ful for detecting early breast cancer, especially
carcinoma in situ.25 Carcinoma in situ can be 1. Jemal A, Murray T, Ward E, et al. Cancer statistics, 2005.
CA Cancer J Clin 2005; 55:10–30.
identified on mammography by the presence of
suspicious microcalcifications. These microcal- 2. Parkin DM, Bray F, Ferlay J, Pisani P. Global cancer statistics,
2002. CA Cancer J Clin 2005; 55:74–108.
cifications are more frequent in intraductal than
in infiltrating cancers,26 which can be easily 3. Tabar L, Yen MF, Vitak B, Chen HH, Smith RA, Duffy SW.
Mammography service screening and mortality in breast
detected by mammography. Because the hard-
cancer patients: 20-year follow-up before and after intro-
ness of carcinoma in situ is lower than that of duction of screening. Lancet 2003; 361:1405–1410.
infiltrating cancer, a little harder than benign tis-
4. Maskarinec G, Meng L, Ursin G. Ethnic differences in mam-
sues, it is not easy to differentiate carcinoma in mographic densities. Int J Epidemiol 2001; 30:959–965.
situ, which our results show. Most false-negative
5. Saarenmaa I, Salminen T, Geiger U, et al. The effect of age
findings on UE were found in early stages of inva- and density of the breast on the sensitivity of breast cancer
sive ductal carcinomas and noninvasive carcino- diagnostic by mammography and ultrasonography. Breast
mas. In the 26 carcinomas misdiagnosed by UE, Cancer Res Treat 2001; 67:117–123.

814 J Ultrasound Med 2007; 26:807–815

26.6.jum.online.q 5/16/07 2:15 PM Page 815

Zhi et al

6. Stavros AT, Thickman D, Rapp CL, Dennis MA, Parker SH, 22. Law J. Cancers detected and induced in mammography
Sisney GA. Solid breast nodules: use of sonography to dis- screening: new screening schedules and younger women
tinguish between benign and malignant lesions. Radiology with family history. Br J Radiol 1997; 70:62–69.
1995; 196:123–134.
23. Moss S, Faulkner K, Law J, Young K. Benefits versus risks
7. Jackson VP. The current role of ultrasonography in breast from mammography: a critical reassessment. Cancer 1997;
imaging. Radiol Clin North Am 1995; 33:1161–1170. 79:628–629.

8. Duncan JL III, Cederbom GJ, Champaign JL, et al. Benign 24. Baker JA, Kornguth PJ, Soo MS, Walsh R, Mengoni P.
diagnosis by image-guided core-needle breast biopsy. Am Sonography of solid breast lesions: observer variability of
Surg 2000; 66:5–10. lesion description and assessment. AJR Am J Roentgenol
1999; 172:1621–1625.
9. Chiou SY, Chou YH, Chiou HJM, et al. Sonographic fea-
tures of nonpalpable breast cancer: a study based on ultra- 25. Benson SR, Blue J, Judd K, Harman JE. Ultrasound is now
sound-guided wire-localized surgical biopsies. Ultrasound better than mammography for the detection of invasive
Med Biol 2006; 32:1299–1306. breast cancer. Am J Surg 2004; 188:381–385.

10. Ophir J, Cespedes I, Ponnekanti H, Yazdi Y, Li X. 26. Ciatto S, Del Turco MR, Bonardi R, et al. Non-palpable
Elastography: a quantitative method for imaging the elas- lesions of the breast detected by mammography: review of
ticity of biological tissues. Ultrason Imaging 1991; 13:111– 1182 consecutive histologically confirmed cases. Eur J
134. Cancer 1994; 30A:40–44.

11. Konig K, Scheipers U, Pesavento A, Lorenz A, Ermert H, 27. Frey H. Real time elastography: a new ultrasound proce-
Senge T. Initial experiences with real-time elastography dure for the reconstruction of tissue elasticity. Radiologe
guided biopsies of the prostate. J Urol 2005; 174:115– 2003; 43:850–855.
117.
28. Kallel F, Ophir J. Limits on the contrast of strain concentra-
12. Garra BS, Cespedes EI, Ophir J, et al. Elastography of breast tions in elastography. Ultrasound Med Biol 1998;
lesions: initial clinical results. Radiology 1997; 202:79–86. 24:1215– 1219.

13. Hiltawsky KM, Kruger M, Starke C, Heuser L, Ermert H,

Jensen A. Freehand ultrasound elastography of breast
lesions: clinical results. Ultrasound Med Biol 2001; 27:
1461–1469.

14. Itoh A, Ueno E, Tohno E, et al. Breast disease: clinical appli-

cation of US elastography for diagnosis. Radiology 2006;
239:341–350.

15. Thomas A, Fischer T, Frey H, et al. Real-time elastography—

an advanced method of ultrasound: first results in 108
patients with breast lesions. Ultrasound Obstet Gynecol
2006; 28:335–340.

16. Lyshchik A, Higashi T, Asato R, et al. Thyroid gland tumor

diagnosis at US elastography. Radiology 2005; 237:
202–211.

17. Schaar JA, De Korte CL, Mastik F, et al. Characterizing vul-

nerable plague features with intravascular elastography.
Circulation 2003; 108:2636–2641.

18. Krouskop TA, Wheeler TM, Kallel F, Garra BS, Hall T. Elastic
moduli of breast and prostate tissues under compression.
Ultrason Imaging 1998; 20:260–274.

19. Cheng SH, Tsou MH, Liu MC, et al. Unique features of
breast cancer in Taiwan. Breast Cancer Res Treat 2000; 63:
213–223.

20. Bird RE, Wallace TW, Yankaskas BC. Analysis of cancers

missed at screening mammography. Radiology 1992;
184:613–617.

21. Boyd NF, Lockwood GA, Martin LJ, et al. Mammographic

densities and risk of breast cancer among subjects with a
family history of this disease. J Natl Cancer Inst 1999; 91:
1404–1408.

J Ultrasound Med 2007; 26:807–815 815