Breast tissue characterization based on ultrasound RF echo modeling and tumor morphology

We propose fractional differencing autoregressive moving average (FARMA) process for modeling RF ultrasound echo. Along with the FARMA model parameters, we use several morphological features extracted from suspected tumors and patient age as potential indicators of malignant or benign masses. Here, we present computerized methods to extract these features and proposed statistical classifiers for breast tissue characterization. We demonstrate the performance of the proposed tissue characterization method on several in vivo ultrasound breast images including benign and malignant tumors. The area of the receiver operator characteristics (ROC) based on 120 in vivo images obtained from 90 patients yields a value of 0.8941, which indicates that proposed tissue characterization method is comparable or better in performance than other methods reported in the literature.