Real-time ultrasound elastography: its potential role in assessment of breast lesions.

We evaluated whether real-time ultrasound elastography (USE) performed in addition to conventional ultrasound (US) can improve the differentiation of benign from malignant breast lesions. Both conventional US and real-time USE were performed in 112 consecutive patients with 139 breast lesions using a Hitachi EUB-8500 US system. Each lesion was assigned an elasticity score according to the degree and distribution of strain induced manually by mild compression. The USE scores (1 to 5) were compared with the BI-RADS assessment categories (1 to 5) obtained with conventional US. Sensitivity, specificity and overall accuracy of each method were determined with surgical pathology as the gold standard. There were 70 benign and 69 malignant lesions. The mean elasticity score was significantly higher for malignant lesions than for benign lesions (4.33 +/- 0.11 vs. 2.10 +/- 0.13, p < 0.01). When a cutoff point of 4 was used, the sensitivity, specificity and accuracy were 85.5, 88.6 and 87% for USE and 94.2, 87.1 and 90.6% for conventional US, respectively. Of the 64 lesions assessed as BI-RADS 2 or 3(i.e., benign) based on conventional US, two were scored as 4 and 5 (i.e., malignant) using USE and were subsequently proven to be malignant. Of the 75 lesions with BI-RADS 4 or 5 category from conventional US, one was scored as a category 1 (benign) with USE and found to be benign by pathology. Our study results suggest that the addition of USE imaging to conventional US could be helpful in the detection and characterization of breast masses.