Differentiation of the breast lesions using statistics of backscattered echoes

The purpose of this study was to evaluate the accuracy of statistical properties of the backscattered ultrasound in differential diagnosis of the breast lesions. The B-mode images, together with the appropriate RF echoes from the breast lesions and surrounding tissues were collected. The RF data was processed for the statistics of the backscattered echo signals, using K and Nakagami distributions characterized by the M and m parameters, respectively. Based on both, M and m parameters, a set of 18 parameters was derived. From the point of view of the sensitivity of detection of cancer, the best score was obtained using maximum value of M parameter, the best specificity was received using the differential Nakagami parameter (the differential values between lesions and surrounding tissues). In conclusion, quantitative sonography is a method which has potential to be a complementary tool for classification of the breast lesions.