Significant differentiation of focal breast lesions: calculation of strain ratio in breast sonoelastography.
ABSTRACT Initial data suggest that elastography can improve the specificity of ultrasound in differentiating benign and malignant breast lesions. The aim of this study was to compare elastography and B-mode ultrasound to determine whether the calculation of strain ratios (SRs) can further improve the differentiation of focal breast lesions.
A total of 227 women with histologically proven focal breast lesions (113 benign, 114 malignant) were included at two German breast centers. The women underwent a standardized ultrasound procedure using a high-end ultrasound system with a 9-MHz broadband linear transducer. B-mode scans and sonoelastograms were analyzed by two experienced readers using the Breast Imaging Reporting and Data System criteria. SRs were calculated from a tumor-adjusted region of interest (mean color pixel density) and a comparable region of interest placed in the lateral fatty tissue. Sensitivity, specificity, and cutoff values were calculated for SRs (receiver-operating characteristic analysis).
The women had a mean age of 54 years (range, 19-87 years). The mean lesion diameter was 1.6 +/- 0.9 cm. Sensitivity and specificity were 96% and 56% for B-mode scanning, 81% and 89% for elastography, and 90% and 89% for SRs. An SR cutoff value of 2.45 (area under the curve, 0.949) allowed significant differentiation (P < .001) of malignant (mean, 5.1 +/- 4.2) and benign (mean, 1.6 +/- 1.0) lesions. The quantitative method of SR calculation was superior to subjective interpretation of sonoelastograms and B-mode scans, with a positive predictive value of 89% compared to 68% and 84% for the other two methods.
Calculation of SRs contributes to the standardization of sonoelastography with high sensitivity and allows significant differentiation of benign and malignant breast lesions with higher specificity compared to B-mode ultrasound but not elastography.
[show abstract] [hide abstract]
ABSTRACT: The aim of this study was to evaluate and compare the diagnostic performance of elastography, B-mode ultrasound (US), and a combination of elastography and B-mode US for the differentiation of small breast masses. A total of 315 breast masses < 1 cm (267 benign, 48 malignant) in 278 patients were examined with B-mode US and elastography. Histopathologic results were used as a reference standard. Two radiologists retrospectively evaluated the B-mode images according to the American College of Radiology Breast Imaging Reporting and Data System and elastographic images according to the elasticity scoring classification system proposed by Itoh et al and the strain ratio. B-mode US and elastography were combined according to the cutoff value. The diagnostic performance of B-mode US, elastography, and the combination of the two modalities was compared using receiver-operating characteristic curve analysis. The mean elasticity score for malignant masses (3.02 ± 1.33) was significantly higher than that for benign masses (1.72 ± 0.78) (P < .001). Areas under the receiver-operating characteristic curves were 0.616 for B-mode US, 0.784 for elasticity score, 0.668 for strain ratio, 0.727 for the combination of B-mode US and elasticity score, and 0.701 for the combination of B-mode US, elasticity score, and strain ratio. The sensitivity, specificity, positive predictive value, and negative predictive value were 93.8%, 51.7%, 25.9%, and 97.9%, respectively, when elasticity score and B-mode US were combined as follows: downgrade of B-mode US assessment category in cases with elasticity scores of 1, no change in cases with scores of 2 or 3, and upgrade in cases with scores of 4 or 5. Elasticity score alone showed the best diagnostic performance, but a combination of B-mode US and elasticity score may have predictive value for the differentiation of benign and malignant lesions <1 cm.Academic radiology 01/2011; 18(1):74-80. · 2.09 Impact Factor