A novel fuzzy logic approach to microcalcification detection

Abstract

Breast cancer continues to be a significant public health problem in the United States. It was estimated that 182,000 new cases of breast cancer would have been diagnosed and 46,000 women have died of breast cancer each year. One out of 8 women will develop cancer at some point during her lifetime in US. Primary prevention seems impossible since the causes of this disease still remain unknown. Early detection is the key to improving breast cancer prognosis. The earlier stage tumors are more easily and less expensively treated. Mammography has been shown to be one of the most reliable methods for early detection of breast carcinomas. However, it is difficult for radiologists to provide both accurate and uniform evaluation for the enormous number of mammograms generated in widespread screening. Because of their clinical relevance, potential subtlety, and the lack of coexisting normal structures having the same appearance, microcalcification clusters (MCCs) are ideal aims for automated detection using image processing and pattern recognition techniques. In this paper, we propose a simple, effective approach to extract the microcalcifications based on fuzzy logic and maximum fuzzy entropy principle. The preliminary results demonstrate the usefulness of the proposed approach for detecting microcalcifications.