COMPUTER-AIDED-DIAGNOSIS FOR IMPROVEMENT OF HEALTH CARE IN DEVELOPING COUNTRIES: COMPUTERIZED DETECTION OF BREAST CANCER

Developing countries have low ratios of radiologists/population and therefore face the challenge of adequate healthcare. We suggest computer-aided-diagnosis as a way of attenuating this problem. Preliminary results of a tool for breast cancer detection by computerized analysis of mammograms are here presented. So far, the tool can detect masses on mammograms, and classify pre-detected microcalcification clusters into benign and malignant types. During testing it yielded a sensitivity of 95.91 % in detecting masses and a good classification rate of 85.29 % for microcalcification clusters