A Novel Computer-Aided Diagnosis System of the Mammograms

Breast cancer is one of the most dangerous tumors for middle-aged and older women in China, and mammography is the most reliable detection method. In order to assist the radiologists in detecting the mammograms, a novel computer-aided diagnosis (CAD) system was proposed in this paper. It carried out a new algorithm using optimal thresholding and Hough transform to suppress the pectoral muscle, applied an adaptive method based on wavelet and filling dilation to extract the microcalcifications (MCs), used a model-based location and segmentation technique to detect the masses, and utilized MLP to classify the MCs and the masses. A high diagnosis precision with a low false positive rate was finally achieved to validate the proposed system.