A Clinical Study Comparing the Diagnostic Performance of Assist Strain Ratio Against Manual Strain Ratio in Ultrasound Breast Elastography.

OBJECTIVE: Strain ratio (SR) is a semiquantitative parameter in differentiating benign from malignant tumors in breast ultrasound elastography. Currently, SR is computed manually and, thus, user dependent. The objective of this study was to evaluate the performance of a new tool assist strain ratio (ASR) and determine how it performs compared with an expert sonologist.

METHODS: Ninety-one patients scheduled for breast biopsy were included in this institutional review board-approved/Health Insurance Portability and Accountability Act-compliant study. For manual strain ratio (MSR), fat and lesion were manually outlined, whereas for ASR, the clinician indicated the lesion center and the fat-to-lesion ratio is computed automatically. Three measurements were obtained for each lesion. The same raw data were used to calculate the MSR and ASR.

RESULTS: The SR thresholds to differentiate benign from malignant tumors were determined using the Youden index. For MSR, the cutoff was 2.7, and for ASR was 2.8. The MSR showed a sensitivity of 88%, specificity of 64%, accuracy of 77%, positive predictive value of 72%, and negative predictive value of 92.1%. Corresponding ASR showed a sensitivity of 86%, specificity of 76%, accuracy of 81%, positive predictive value of 79%, and negative predictive value of 84%. The areas under the curve for the MSR and ASR were 0.83 and 0.85, respectively. The average coefficients of variation for the MSR and ASR measurements were 43% and 30%, respectively.

CONCLUSION: Assist strain ratio demonstrated similar diagnostic performance compared with MSR. In addition, the coefficient of variation of ASR is lower, implying lower intraoperator dependency. Thus, ASR may aid less-experienced scanners in obtaining improved results.