LAUSR

Breast cancer is one of the main cancers that effect of the women’s health. This cancer is one of the most important health issues in the world and because of that, diagnosis in the beginning and appropriate cure is very effective in the recovery and survival of patients, so image processing as a decision-making tool can assist physicians in the early diagnosis of cancer. Image processing mechanisms are simple and non-invasive methods for identifying cancer cells that accelerate early detection and ultimately increase the chances of cancer patients surviving. In this study, a pipeline methodology is proposed for optimal diagnosis of the breast cancer area in the mammography images. Based on the proposed method, after image preprocessing and filtering for noise reduction, a simple and fast tumors mass segmentation based on Otsu threshold segmentation and mathematical morphology is proposed. Afterward, for simplifying the final diagnosis, a feature extraction based on 22 structural features is utilized. To reduce and pruning the useless features, an optimized feature selection based on a new developed design of Water Strider Algorithm (WSA), called Guided WSA (GWSA). Finally, the features injected to an optimized SVM classifier based on GWSA for optimal cancer diagnosis. Simulations of the suggested method are applied to the DDSM database. A comparison of the results with several latest approaches are performed to indicate the method higher effectiveness.