Automated Mass Detection System in Mammograms

Breast cancer remains as a leading cause of cancer deaths among women in many parts of the world. Cancer is an abnormal, continual multiplying of cells. Most types of cancer cells eventually form a lump or mass called a tumor. In this paper, an automatic CADe system for mammographic mass detection that uses complex texture features is employed. The presence of ambiguous margins of lesions and visual fatigue causes radiologists to miss roughly 10-30% of tumor in breast cancer screening. Texture, the pattern of information or arrangement of the structure found in an image, is an important feature in image processing. It also refers a method for the feature extraction of mammograms using Grey Level Co-occurrence Matrix (GLCM) and Optical Density Co-occurrence Matrix (ODCM) for suspicious areas of breast to identify tumor. Thus the radiologists are able to segment masses on mammogram which are surrounded by the perplexed tissues. Index Terms — Computer-aided detection (CADe) system, mammographic mass detection, optical density