Recognition of Breast Abnormalities Using Phase Features

Early detection of breast pathologies and proper treatment increase the likelihood of a cure, and, as a result, life expectancy. Currently, methods and algorithms for computer aided detection (CAD) systems are being actively developed. The traditional approach to designing such systems consists in selecting and calculating the features of the region of interest from the source data, followed by the selection of a model for their classification using machine learning methods. This paper proposes a method for detecting and classifying breast anomalies based on local energy and phase congruency and a controlled machine learning classifier. Experimental results are presented using a digital mammography dataset and evaluated using various performance criteria.