ABSTRACT
PURPOSE: We compared elastography, B-mode ultrasound and mammography to determine whether raw data calculation of strain ratios (SRs) can further improve the differentiation of focal breast lesions. MATERIALS AND METHODS: 201 women with histologically proven focal breast lesions (85 benign, 116 malignant) were included at two German breast centers. Patients underwent a standardized ultrasound procedure using high-end ultrasound system with a 9-MHz broadband linear transducer. Two experienced readers analyzed the B-mode scans and mammograms using the BI-RADS criteria, while elastograms were analyzed using the Tsukuba score. SRs were calculated from a tumor-adjusted ROI and a comparable ROI placed in the lateral fatty tissue. The sensitivity, specificity, and cutoff values were calculated for SRs (ROC analysis). RESULTS: The median age was 53 years. The sensitivity and specificity were 85 %/ 60 % for B-mode scanning, 85 %/ 68 % for elastography, 78 %/ 62 % for mammography, and 95 %/ 74 % for SRs. An SR cutoff value of 2.27 (AUC 0.907) allowed significant differentiation (p < 0.001) between malignant and benign lesions. The quantitative SR calculation was superior to subjective interpretation of B-mode scans and sonoelastograms with a positive predictive value of 83 % versus 78 % and 74 %, and equal to mammograms. CONCLUSION: Strain ratio calculation contributes to the standardization of sonoelastography with high sensitivity and allows significant differentiation between benign and malignant breast lesions with a higher specificity compared to B-mode, subjective evaluation of elastography and mammography.
