A METHOD FOR MICROCALISIFICATIONS DIAGNOSIS IN BREAST MAMMOGRAMS

Abstract

Breast cancer is the main cause of death for women above age 40. In this paper a method is proposed to develop a Computer-Aided Diagnosis (CADx), this method provide a second opinion in microcalcifications diagnostic and making decisions (classify microcalcifications as benign or malignant). The proposed method for microcalcifications diagnosis splits into three steps: The first step extracts the region of interest (ROI). The second step is the features extraction, where we used a set of features from (ROI) by applying wavelet decomposition. The third step is the classification process where discrimination between benign and malignant is performed using a Nearest Neighbor Classifier. The proposed method was evaluated using the Mammographic Image Analysis Society (MIAS) mammographic databases. The proposed method has achieved satisfactory results.