Breast Cancer Research and Treatment

, Volume 129, Issue 1, pp 89–97 | Cite as

Clinical application of shear wave elastography (SWE) in the diagnosis of benign and malignant breast diseases

  • Jung Min Chang
  • Woo Kyung Moon
  • Nariya Cho
  • Ann Yi
  • Hye Ryoung Koo
  • Wonsik Han
  • Dong-Young Noh
  • Hyeong-Gon Moon
  • Seung Ja Kim
Clinical trial

Abstract

Shear wave elastography (SWE) is an emerging technique which can obtain quantitative elasticity values in breast disease. We therefore evaluated the diagnostic performance of SWE for the differentiation of breast masses compared with conventional ultrasound (US). Conventional US and SWE were performed by three experienced radiologists for 158 consecutive women who had been scheduled for US-guided core biopsy or surgical excision in 182 breast masses (89 malignancies and 93 benign; mean size, 1.76 cm). For each lesion, quantitative elasticity was measured in terms of the Young’s modulus (in kilopascals, kPa) with SWE, and BI-RADS final categories were assessed with conventional US. The mean elasticity values were significantly higher in malignant masses (153.3 kPa ± 58.1) than in benign masses (46.1 kPa ± 42.9), (P < 0.0001). The average mean elasticity values of invasive ductal (157.5 ± 57.07) or invasive lobular (169.5 ± 61.06) carcinomas were higher than those of ductal carcinoma in situ (117.8 kPa ± 54.72). The average mean value was 49.58 ± 43.51 for fibroadenoma, 35.3 ± 31.2 for fibrocystic changes, 69.5 ± 63.2 for intraductal papilloma, and 149.5 ± 132.4 for adenosis or stromal fibrosis. The optimal cut-off value, yielding the maximal sum of sensitivity and specificity, was 80.17 kPa, and the sensitivity and specificity of SWE were 88.8% (79 of 89) and 84.9% (79 of 93). The area under the ROC curve (Az value) was 0.898 for conventional US, 0.932 for SWE, and 0.982 for combined data. In conclusion, there were significant differences in the elasticity values of benign and malignant masses as well as invasive and intraductal cancers with SWE. Our results suggest that SWE has the potential to aid in the differentiation of benign and malignant breast lesions.

Keywords

Breast ultrasound Shear wave elastography Breast masses Elasticity values 

Introduction

Ultrasound (US) elastography is a tool that reflects the hardness of a lesion, and a recent study using the currently available US technology has shown that elastography has nearly the same diagnostic performance as conventional US, with 86.5% (45 of 52) sensitivity, 89.8% (53 of 59) specificity, and 88.3% (98 of 111) accuracy in the differentiation of benign from malignant solid breast masses [1].

However, US elastography with freehand compression has several disadvantages in that the elasticity map obtained is highly dependent on the organ’s compressibility limits under stress and on the extent of tissue compression applied. Thus, the acquired information is not quantitative, but is rather a measurement of local strain estimated at a given location in the tissue, which depends on the surrounding mechanical properties. In addition, from the user’s perspective, the reliability of the imaging technique depends on the skill of the operator to correctly compress the tissue, resulting in a substantial amount of interobserver variability during data acquisition and interpretation [2, 3, 4].

To overcome this problem, the quantitative elastography technique of shear wave elastography (SWE) has been developed. This new technique depends less on the individual operator, is reproducible, and is quantitative [5]. It works by remotely inducing mechanical vibrations through acoustic radiation force created by a focused US beam. The displacement induced at the focus generates a shear wave that conveys information linked to the local viscoelastic properties of the tissue, thus enabling a quantitative approach to elasticity values. A very fast (5000 frames per second) US acquisition sequence is then used to capture the propagation of the shear waves [6, 7, 8, 9]. Recently, several studies on quantitative SWE have been published [5, 6, 7, 8, 9]. They have shown that thus far the addition of shear imaging has improved the positive predictive values of US. However, these studies included only a small number of cases and have presented limited SWE data based on histologic findings.

Therefore, we undertook this current study to prospectively assess the quantitative values of various breast lesions in a larger population and to evaluate the diagnostic performance of SWE in the differentiation of breast masses compared with conventional US, using histological analysis as the standard of reference.

Patients and methods

Patients and breast masses

This prospective study was conducted with institutional review board (IRB) approval, and informed consent was provided by all patients. Between March 2010 and May 2010, 186 breast masses in 162 consecutive women who had been scheduled to undergo US-guided percutaneous needle biopsy or surgical excision were examined with SWE. Two patients who underwent neoadjuvant chemotherapy were excluded, and two patients with multifocal breast cancer were excluded because of difficulty in correlating the US visible lesion and the pathologic results due to multiplicity. Finally, 182 breast masses in 158 consecutive women (mean age, 48.1 years; age range, 22–79 years) constituted our study population. Of these 158 women, 98 women were asymptomatic, 53 women presented with a palpable mass, and seven showed nipple discharge. Mammograms were available for 160 masses in 139 women during the US examinations. Lesions were observed as a mass in 47 cases (29%), as a mass with microcalcifications in 27 (17%), as an architectural distortion in five (3%), and as a focal asymmetry in 21 cases (13%). No mammographic abnormalities were found for 60 (37%) lesions. According to the American College of Radiology (ACR) Breast Imaging Reporting and Data System (BI-RADS) [10], the final assessments of the 182 solid breast masses determined prior to biopsy were as follows: category 3 (probably benign) for 16 masses, category 4 (suspicious abnormality) for 99, and category 5 (highly suggestive of malignancy) for 67. Masses were confirmed through US-guided core needle biopsy (14-gauge automated gun or 11-gauge vacuum-assisted) (n = 96) or surgical excision (n = 86). Of the 182 masses, 89 (49%) were malignant and 93 (51%) were benign. Malignant masses included infiltrating ductal carcinomas (n = 72), infiltrating lobular carcinomas (n = 7), ductal carcinoma in situ (DCIS) (n = 7), a mucinous cancer (n = 1), an apocrine cancer (n = 1), and a metaplastic cancer (n = 1). The benign lesions were found to consist of 33 fibroadenomas, seven papillomas, two adenoses, one stromal fibrosis, and 50 fibrocystic changes. The mean size of the malignant masses was 2.4 ± 1.19 cm (range 0.5–6.5 cm) and that of the benign masses was 1.1 ± 0.60 cm (range 0.4–2.9 cm). All benign lesions remained unchanged during the follow-up period (mean 10.7 months, range 10–12 months), and mastectomy or breast conserving surgery was performed for all breast cancers

US examinations and image evaluation

Conventional US and SWE were obtained using the Aixplorer® US system (SuperSonic Imagine, Aix en Provence, France), by one of three radiologists (J.M.C., N.C., and W.K.M.) with 7–20 years of experience in breast US and 12–36 months of experience in US elastography, with knowledge of the clinical and mammographic findings. Lesion location and size were recorded. BI-RADS scores were prospectively assessed by the radiologists who had performed the B-mode US.

After conventional US, SWE images were generated by the same radiologists. The lesions were located in the center of the elasticity boxes, and quantitative elasticity values were measured by taking the mean value and standard deviations of the region of interest (ROI) boxes centered on the elasticity image at the location of maximum stiffness. The size of each ROI box was 2.5 × 1.5 cm by default, with a maximal size of 3 × 2.5 cm. The mean values were recorded in all cases. The data acquisition procedure took approximately 2–3 min per case.

Data collection and statistical analysis

The mean elasticity values of all benign and malignant pathologies were recorded, and the average of the mean values for each subcategorized pathology was assessed. The average mean elasticity values were also compared between benign and malignant masses and invasive cancer and DCIS using the nonparametric Mann–Whitney test. Two-tailed P values of less than 0.05 were considered to indicate statistical significance.

Receiver operating characteristic (ROC) curves for the B-mode US and SWE images were analyzed to evaluate diagnostic performance and to compare the performance of established BI-RADS categories and elasticity values. The optimal cut-off values for B-mode and SWE, yielding the maximal sum of sensitivity and specificity, were calculated. The sensitivity, specificity, and positive (PPV) and negative predictive values (NPV) using the cut-off values were measured. Logistic regression model was used to perform ROC analysis of combined data of both B-mode and SWE images. Since no malignant lesions were observed in category 3 and no benign lesions were observed in category 5 on conventional B-mode US, which is known as ‘separation problem’, the logistic regression was fitted using penalized maximum likelihood [11]. To summarize overall performances, areas under the ROC curves (Az) were calculated and compared between the these techniques using SAS 9.1. Since some patients have more than one breast mass, cluster effect was considered in ROC curve analysis.

Statistically significant differences between Az values were reported at 95% confidence intervals. The mean differences were regarded as being statistically significant at the 5% level when the corresponding confidence interval did not encompass zero.

Results

Elasticity values of the solid masses

Malignant masses had an average mean elasticity of 153.3 kPa ± 58.1, whereas benign masses demonstrated an average mean value of 46.1 kPa ± 42.9 (Table 1) (Fig. 1). The difference was statistically significant (P < 0.0001). The average mean elasticity was 157.5 kPa ± 57.07 for invasive ductal cancer (Fig. 2), 169.5 kPa ± 61.06 for invasive lobular cancer, 190.9 kPa for mucinous cancer, 165.3 kPa for metaplastic cancer, and 60.9 kPa for apocrine cancer. In contrast, DCIS (Fig. 3) showed lower elasticity values than other invasive cancers, with an average mean value of 117.8 kPa ± 54.72. The difference between the invasive cancers and DCIS was statistically significant (P = 0.049).
Table 1

Histopathologic types of malignant and benign lesions and the average mean elasticity values

Histology

N

Average mean kPa (± standard deviation)

Range

Malignant

89

153.3 ± 58.1

46.95–300.0

 Ductal carcinoma in situ

7

117.75 ± 54.72

46.95–193.3

  Non-high

4

114.68 ± 62.17

59.14–148.2

  High

3

109.49 ± 75.49

46.95–193.3

 Invasive ductal cancer (NOS type)

72

157.5 ± 57.07

58.34–300

  Grade I

3

138.19 ± 28.76

60.9–165.79

  Grade II

30

144.58 ± 58.03

58.34–231.85

  Grade III

39

167.70 ± 56.96

65.95–300.0

 Invasive lobular cancer

7

169.5 ± 61.06

107.63–283.84

 Mucinous cancer

1

190.9

190.9

 Metaplastic cancer

1

165.3

165.3

 Apocrine cancer

1

60.9

60.9

Benign

93

46.07 ± 42.9

3.3–192.51

 Fibroadenoma

33

49.58 ± 43.51

5.89–192.51

 Fibrocystic change

50

35.3 ± 31.2

3.3–149.3

 Intraductal papilloma

7

69.5 ± 63.2

10.3–172.43

 Adenosis

2

123.9 ± 62.3

79.9–168.0

 Stromal fibrosis

1

149.5

 
Fig. 1

Box-and-whisker plot of the Young’s modulus was estimated using SWE. The elasticity values were significantly different between malignant and benign lesions (P < 0.001, the Mann–Whitney test). Boxes values from lower to upper quartiles, central lines medians; whiskers extend from minimal to maximal values. Dots outliers

Fig. 2

SWE and B-mode images of a 1.7-cm grade 3 invasive ductal carcinoma in a 49-year-old female showing high stiffness (mean 273.8 kPa). This mass was categorized as a BI-RADS category 5 lesion

Fig. 3

SWE and B-mode images of a 1.3-cm low grade DCIS in a 68-year-old female showing a suspicious B-mode feature, BI-RADS category 5, but relatively low stiffness (mean 59.1 kPa)

The distribution of average mean elasticity values for benign lesions according to their histologic findings is listed in Table 1. The average mean values were 49.58 kPa ± 43.51 for fibroadenoma and 35.3 kPa ± 31.2 for fibrocystic changes. In benign cases, most lesions showed mean values lower than 50 kPa (Fig. 4). However, the average mean value for intraductal papilloma was 69.5 kPa ± 63.2 (Fig. 5), and the average mean value for adenosis or stromal fibrosis was 149.5 kPa ± 132.4, as high as the values seen in the malignant masses.
Fig. 4

SWE and B-mode images of a benign fibroadenoma in a 61-year-old female presenting with benign elastography features (mean 12.9 kPa)

Fig. 5

SWE and B-mode images of an intraductal papilloma in a 30-year-old female showing high stiffness (mean 172.4 kPa). This mass was categorized as a BI-RADS category 4 lesion

The mean elasticity values according to lesion size seen in the B-mode images are shown in Fig. 6. The quantitative elasticity values were higher in larger masses for both the benign and malignant groups. However, for each lesion size category, the mean value was still significantly higher for malignant lesions than for benign lesions (P < 0.001), and even in the 4-5-mm-sized nodules, definite differentiation was noted for the elasticity values between benign and malignant nodules (Fig. 7).
Fig. 6

Mean elasticity values according to lesion size in B-mode images. The numbers on the graphs are the mean kPa values of both malignant and benign lesions according to their size. For each lesion size category, the mean value was still significantly higher for malignant lesions than for benign lesions (P < 0.001)

Fig. 7

SWE and B-mode images of a 0.5 cm invasive ductal carcinoma in a 49-year-old female showing high stiffness (mean 86.2 kPa). This mass was categorized as a BI-RADS category 4 lesion

Sensitivity, specificity, and ROC analysis

The diagnostic performance of conventional US and that of SWE at various cut-off points is shown in Table 2. For SWE, the optimal cut-off value, yielding the maximal sum of sensitivity and specificity, was 80.17 kPa, and the sensitivity, specificity, PPV and NPV of the SWE were 88.8% (79 of 89), 84.9% (79 of 93), 84.9% (79 of 94), and 88.8% (79 of 89). Other values of 50 and 100 kPa were also tested to compare the diagnostic performance. When a cut-off point between categories 3 and 4 was used, conventional US had 100% (89 of 89) sensitivity, 17.2% (16 of 93) specificity, a 53.6% (89 of 166) PPV, and a 100% (16 of 16) NPV. When a cut-off point between categories 4 and 5 was used, conventional US had 75.3% (67 of 89) sensitivity, 100% (93 of 93) specificity, a 100% (67 of 67) PPV, and 80.9% (93 of 115) NPV.
Table 2

Diagnostic performances of SWE and conventional US at various cutoff points for the diagnosis of benign and malignant lesions

 

Sensitivity (%)

Specificity (%)

Accuracy (%)

Elasticity value (kPa)

 50.00

98.9% (88/89)

66.7% (62/93)

82.4% (150/182)

 80.17

88.8% (79/89)

84.9% (79/93)

86.8% (158/182)

 100.0

80.9% (72/89)

88.2% (82/93)

84.6% (154/182)

Conventional US categorya

 Between 1 and 2

100% (89/89)

0% (0/93)

48.9% (89/182)

 Between 2 and 3

100% (89/89)

0% (0/93)

48.9% (89/182)

 Between 3 and 4

100% (89/89)

17.2% (16/93)

57.9% (106/182)

 Between 4 and 5

75.3% (67/89)

100% (93/93)

87.9% (160/182)

Numbers in parentheses were used to calculate percentages

aConventional US category was defined according to the BI-RADS classification for US

The Az value was 0.898 ± 0.02 (95% confidence interval: 0.859, 0.936) for conventional US and 0.932 ± 0.02 (95% confidence interval: 0.896, 0.967) for SWE, which was not significantly different (P = 0.185). The combination of conventional US and SWE showed the highest discriminating power for the detection of a malignancy, and the Az value was 0.982 ± 0.007 (95% CI: 0.968–0.995), significantly high than both conventional US (P < 0.0001) and SWE (P = 0.0004) (Fig. 8).

False-negative lesions in SWE

Among the malignant masses, 10 of 89 lesions (11.2%) showed values lower than 80.17 kPa, ranging from 46.95 to 70.34 kPa. The mean size of these false-negative lesions was 2.22 cm ± 0.98. Four of them were initially categorized as BI-RADS category 4, and six lesions were classified as BI-RADS category 5. Six of these lesions were invasive ductal carcinoma, three were DCIS, and one was apocrine cancer.

False-positive lesions in SWE

There were 14 of 93 benign lesions (15.1%) that showed values higher than 80.17 kPa, ranging from 86.56 to 192.5 kPa. The mean size of these false-positive lesions was 1.69 cm ± 0.82. Only three of them were initially categorized as BI-RADS category 3. The remaining 11 were initially categorized as BI-RADS category 4, which means 78.6% of these lesions showed suspicious findings in both B-mode images and SWE.

Discussion

Our study investigated the quantitative elasticity values of various breast lesions using SWE and evaluated its diagnostic performance for the differentiation of benign and malignant breast masses. We found that the mean elasticity values were significantly higher in malignant masses (153.3 kPa ± 58.1) than in benign masses (46.1 kPa ± 42.9), (P < 0.0001). These significant differences were noted in all subcategorized groups classified by lesion size. Even in 0.5-cm cancers, the elasticity values for malignancies were significantly higher than benign nodules with the same size. We also found that the areas under the ROC curves (Az values) were 0.898 for conventional US and 0.932 for SWE (P = 0.250), and the best performance was noted for the combined data of conventional US and SWE (Az value of 0.982). Our findings are concordant with other studies using SWE, and the mean values of the breast masses are quite similar to the findings of Athanasiou et al. [5]. The mean elasticity value of benign masses, mainly consisting of fibroadenomas, aberrations of normal development and involution, and cysts, was 45.3 kPa ± 41.1. The mean elasticity value of malignant masses was 146.6 kPa ± 40.05 in their study. In addition, in the report by Evans et al. [12], the average mean value of 28 invasive cancers was 140 kPa, and the average mean value of fibroadenoma was 28 kPa.
Fig. 8

ROC curve to compare the performance of BI-RADS categorization in B-mode, SWE mean stiffness, and combined data

Even though significant differences were noted between benign and malignant breast lesions in general, an overlap of elasticity between benign and malignant lesions in the breast can interfere with the differentiation of malignancies from benign lesions. In our study, the elasticity range of both benign and malignant pathologies showed a substantial amount of overlap. Early-stage breast cancers, malignancies that contain large areas of necrosis [13], and specific tumor types such as mucinous carcinoma were reported to be the causes of benign-appearing elastography [13, 14]. In our larger study population, there were seven DCIS, seven ILC, one mucinous cancer, one metaplastic cancer, and one apocrine cancer. A high kPa value was noted in the 0.5-cm mucinous cancer, which was different from our expectation that mucinous cancer might not be stiff. DCIS and apocrine cancer showed lower kPa values than other invasive cancers. In contrast with our larger population, no DCIS was included in the study reported by Athanasiou et al. [5], and only one DCIS was included in the report by Evans et al. [12]. The mean elasticity value of the lesion was 76 kPa; this value is in the range of our DCIS cases. Inclusion of DCIS or variable stages of invasive cancer in the malignant group will make elasticity values more heterogeneous and thus will produce overlap with the benign lesions.

Apart from the specific histologic type of tumor, the size of the tumor can also affect elasticity values. According to Evans et al. [12], invasive cancers with an US size of <15 mm had an average mean elasticity of 109 kPa, compared to an average value for lesions ≥ 15 mm of 167 kPa. Using static elastography, higher mean scores were noted in the larger lesions for both malignant and benign lesions [1]. Small cancers were not as stiff as larger cancers; however, this does not reduce the stiffness values to the benign range. In our study, the mean values for different lesion sizes show higher values in the larger masses in both benign and malignant lesions, even though statistical significance was not tested. Not only size, but also the depth of the lesion and breast thickness may affect the measurement of mean values [13, 15].

In previous reports of both static and SWE, many authors concluded that elastography did not have the potential to replace conventional B-mode US for the detection of breast cancer; however, they found it may be useful as a complement to conventional US to improve diagnostic performance [1, 13, 14, 16, 17, 18]. In addition, overall diagnostic performance was increased with the combination of conventional US and SWE [5]. In our study, the ROC analysis using conventional B-mode without and with elasticity data was performed and the significant improvement was found when conventional B-mode was used with elasticity data.

According to Athanasiou et al., adding the elasticity value modified the BI-RADS 4 score to BI-RADS 3 for 13 lesions, meaning that biopsy could have been avoided in 46% of cases, and a short 6-month follow-up could have been completed [5]. In our study, there were 77 BI-RADS category 4 benign lesions, and among those lesions 66 masses showed an elasticity value equal to or lower than 80.17 kPa. Using this cut-off value, we can eliminate 85.7% of false-positive biopsies.

However, four BI-RADS category 4 malignant masses would also be downgraded and potentially missed by the application of this cut-off point. These were a 1-cm invasive cancer, one low grade and one high grade DCIS, and a 0.9-cm apocrine cancer that showed low elasticity values. In addition, there were still 14 benign lesions that showed elasticity values higher than 80.17.

The cut-off value of 50 kPa was selected in previous research by Evans et al. [12]. With this cut-off point, the diagnostic performance of SWE was significantly superior to B-mode imaging, showing 97% sensitivity, 83% specificity, 88% PPV, and 95% NPV. Only one DCIS was included in that study, and the lesion with a mean stiffness below the 50 kPa threshold was a 1.2-cm grade 2 invasive ductal carcinoma, which was classified as BI-RADS 3. If we apply the same cut-off point of 50 kPa, although we have benign pathologies of notably higher elasticity values, our diagnostic performance becomes 98.9% (88/89) sensitivity, 66.7% (62/93) specificity, 74% (88/119) PPV, and 98% (62/63) NPV; we now have higher values of sensitivity and NPV with lower values of specificity and PPV. The choice of cut-off is dependent on what use will be made of the data. The application of a certain cut-off value is not a simple task, and a more heterogeneous study population will make the decision of an appropriate cut-off value more complicated. Thus, a meticulous investigation should be carried out to find the best cut-off point to optimize the differentiation of benign and malignant lesions.

This study had some limitations. First, we did not include cystic lesions in our study population. SWE is reported to be useful in demonstrating and characterizing cystic lesions where, regardless of a typical or complicated B-mode US appearance, an absence of signal is noted due to the absence of any shear wave propagation in liquid areas. We did not note any absence of signal in our study population. Second, multivariate analysis for the evaluation of confounding factors was not performed in our study. Elasticity values according to histologic differentiation were analyzed; however, individual lesion size, histologic grade, the surrounding fibrotic component from pathology, and the existence of microcalcifications were not statistically analyzed. Lesion depth and breast thickness which are known to be able to influence diagnostic performance [19] were also not considered. Thus, an additional study is necessary to confirm the correlation of elasticity values with each variable, and to find out the variables which can affect the false-positive or false-negative results. Third, even though, the combined diagnostic performance of B-mode imaging and SWE was evaluated, these results are limited to our study, and a specific guideline for the combination of BI-RADS score and elasticity values was not provided. In clinical practice, most radiologists consider both B-mode and elastography findings for a clinical decision. More clinical studies on how best to combine SWE and B-mode US findings to enhance benign and malignant differentiation are necessary.

In conclusion, SWE showed significant differences in mean elasticity values for benign and malignant masses. The combination of conventional B-mode US findings and elasticity values could increase the diagnostic performance of breast US, and if validated by larger studies including a greater variety of breast pathologies, SWE combined with B-mode characterization may potentially increase our level of confidence regarding the final assessment of various breast lesions.

Notes

Acknowledgment

This study was supported by a grant from the Innovative Research Institute for Cell Therapy (A062260) and by a grant from the National R&D Program for Cancer Control (A01185), Ministry of Health & Welfare, Republic of Korea. The authors appreciated the statistical advice from the Medical Research Collaborating Center at the Seoul National University Hospital and the Seoul National University College of Medicine.

Conflict of interest

Authors declare no conflict of interest.

References

  1. 1.
    Itoh A, Ueno E, Tohno E et al (2006) Breast disease: clinical application of US elastography for diagnosis. Radiology 239:341–350PubMedCrossRefGoogle Scholar
  2. 2.
    Fleury EdFC, Fleury JC, Piato S, Roveda D Jr (2009) New elastographic classification of breast lesions during and after compression. Diagn Interv Radiol 15:96–103PubMedGoogle Scholar
  3. 3.
    Burnside ES, Hall TJ et al (2007) Differentiating benign from malignant solid breast masses with US strain imaging. Radiology 245:401–410PubMedCrossRefGoogle Scholar
  4. 4.
    Regner DM, Hesley GK, Hangiandreou NJ et al (2006) Breast lesions: evaluation with US strain imaging-clinical experience of multiple observers. Radiology 238:425–437PubMedCrossRefGoogle Scholar
  5. 5.
    Athanasiou A, Tardivon A, Tanter M et al (2010) Breast lesions: quantitative elastography with supersonic shear imaging—preliminary results. Radiology 256:297–303PubMedCrossRefGoogle Scholar
  6. 6.
    Bercoff J, Tanter M, Fink M (2004) Supersonic shear imaging: a new technique for soft tissue elasticity mapping. IEEE Trans Ultrason Ferroelectr Freq Control 51:396–409PubMedCrossRefGoogle Scholar
  7. 7.
    Bercoff J, Tanter M, Muller M, Fink M (2004) The role of viscosity in the impulse diffraction field of elastic waves induced by the acoustic radiation force. IEEE Trans Ultrason Ferroelectr Freq Control 51:1523–1536PubMedCrossRefGoogle Scholar
  8. 8.
    Bercoff J, Chaffai S, Tanter M et al (2003) In vivo breast tumor detection using transient elastography. Ultrasound Med Biol 29:1387–1396PubMedCrossRefGoogle Scholar
  9. 9.
    Tanter M, Bercoff J, Athanasiou A et al (2008) Quantitative assessment of breast lesion viscoelasticity: initial clinical results using supersonic shear imaging. Ultrasound Med Biol 34:1373–1386PubMedCrossRefGoogle Scholar
  10. 10.
    Mendelson EB, Baum JK, Berg WA, Merritt CRB, Rubin E (2003) Breast Imaging Reporting and Data System, BI-RADS: Ultrasound, 1st edn. American College of Radiology, RestonGoogle Scholar
  11. 11.
    Heinze G (2004) A SAS macro, S-PLUS library and R package to perform logistic regression without convergence problems. Technical Report 2/2004: Medical University of Vienna, ViennaGoogle Scholar
  12. 12.
    Evans A, Whelehan P, Thomson K et al (2010) Quantitative shear wave ultrasound elastography: initial experience in solid breast masses. Breast Cancer Res 12:R104PubMedCrossRefGoogle Scholar
  13. 13.
    Raza S, Odulate A, Ong EM, Chikarmane S, Harston CW (2010) Using real-time tissue elastography for breast lesion evaluation: our initial experience. J Ultrasound Med 29:551–563PubMedGoogle Scholar
  14. 14.
    Scaperrotta G, Ferranti C, Costa C et al (2008) Role of sonoelastography in non-palpable breast lesions. Eur Radiol 18:2381–2389PubMedCrossRefGoogle Scholar
  15. 15.
    Thomas A, Fischer T, Frey H et al (2006) Real-time elastography—an advanced method of ultrasound: first results in 108 patients with breast lesions. Ultrasound Obstet Gynecol 28:335–340PubMedCrossRefGoogle Scholar
  16. 16.
    Cho N, Moon WK, Park JS, Cha JH, Jang M, Seong MH (2008) Nonpalpable breast masses: evaluation by US elastography. Korean J Radiol 9:111–118PubMedCrossRefGoogle Scholar
  17. 17.
    Egorov V, Kearney T, Pollak SB et al (2009) Differentiation of benign and malignant breast lesions by mechanical imaging. Breast Cancer Res Treat 118:67–80PubMedCrossRefGoogle Scholar
  18. 18.
    Chung SY, Moon WK, Choi JW, Cho N, Jang M, Kim KG (2010) Differentiation of benign from malignant nonpalpable breast masses: a comparison of computer-assisted quantification and visual assessment of lesion stiffness with the use of sonographic elastography. Acta Radiol 51:9–14PubMedCrossRefGoogle Scholar
  19. 19.
    Chang JM, Moon WK, Cho N, Kim SJ (2011) Breast mass evaluation: factors influencing the quality of US elastography. Radiology 259:59–64PubMedCrossRefGoogle Scholar

Copyright information

© Springer Science+Business Media, LLC. 2011

Authors and Affiliations

  • Jung Min Chang
    • 1
  • Woo Kyung Moon
    • 1
  • Nariya Cho
    • 1
  • Ann Yi
    • 1
  • Hye Ryoung Koo
    • 1
  • Wonsik Han
    • 2
  • Dong-Young Noh
    • 2
  • Hyeong-Gon Moon
    • 2
  • Seung Ja Kim
    • 3
  1. 1.Department of Radiology and Clinical Research InstituteSeoul National University Hospital and the Institute of Radiation Medicine, Seoul National University Medical Research CenterSeoulKorea
  2. 2.Department of SurgerySeoul National University HospitalSeoulKorea
  3. 3.Department of RadiologySeoul Metropolitan Government Seoul National University, Boramae Medical CenterSeoulKorea

Personalised recommendations