A guided ant colony algorithm for detection of microcalcification clusters

Women suffering from breast cancer have been taken as a severe concern all around the world, as it directly distresses the next generation to come. The Microcalcification Cluster in the mammograms is one of the vital sign for the early diagnosis of breast cancer. In this paper, an improved Ant Colony Algorithm is proposed for the segmentation of MCs in two-steps: (a) watershed segmentation is applied on the edge map of the digital mammograms, which reduces the over-segmentation property of watershed algorithm enormously, and (b) a Guided Ant Colony Algorithm (GACA) is applied to merge the segmented regions together based on their homogeneity to produce the final segmentation into two different clusters such as normal and abnormal tissues. The performance of the classification is evaluated with Receiver Operating Characteristic curve analysis. The greatest accuracy of 96% is achieved by GACA states the better performance of the proposed method.