Fusion of Two View Mammographic Texture Information Through Binary BAT Algorithm for Enhanced Breast Cancer Detection

Female breast cancer is a mortal disease. Every year its incidences and mortality rate increase globally. According to statistics studies, number of breast cancers estimated during 2010, 2015 and 2020 are 0.09, 1.06 and 0.123 million, respectively. Symptoms are not obvious in the beginning stage. Therefore, the early detection is difficult. Mammography is a gold standard for more than forty years and Cranio-Caudal and Medio-Lateral Oblique views are normally used in diagnosing breast diseases. These two views are results of scanning the breast in different angle. Thus, they provide some unique information in addition to the information common to both views. Hence combining the information of these two views for diagnosis may produce better results. To improve the detection performance, fusion of texture information from the images of these two views through Binary BAT algorithm is suggested in this paper. An improved performance metric was obtained using the proposed method.