Main Article Content
In this research, a hybrid segmentation technique has been put forward to identify suspicious regions on digital mammograms based on Expectation Information Measure. The proposed algorithm measures different forms of mammograms (fatty, fatty-glandular and dense glandular) on the images from the mini-MIAS (Mammogram Image Analysis Society database)). To exhibit their efficacy, the suggested approach is contrasted with various segmentation strategies such as Expectation Maximization and Maximum Entropy Threshold. In the identification of suspect areas, the experimental findings on mammography images showed efficacy. This research can be part of the creation of a decision that is computer-aided (CAD) diagnosis to detect breast cancer early on.