Multi-mass breast cancer classification based on hybrid descriptors and memetic meta-heuristic learning

摘要: Among the causes of death in world, breast cancer is considered most common cause mortality among women to extent that one five deaths attributed incidence this cancer. In paper, we introduce a computer-aided detection approach multiple classifications masses. We tried separate and intelligently recognize different masses by means mammograms so first step, with pre-processing, pectoral region segmented from other parts areas are primarily clustered K-means method. next using aggregation efficient features such as texture features, Pseudo–Zernike moments, wavelet will be extracted input image simulated annealing algorithm reduce size feature vector. The final step classification possible mammogram assessment its severity based on memetic meta-heuristic adaptive neuro-based fuzzy inference system which optimizer shuffled frog-leaping algorithm. proposed method evaluated 322 images taken Mini-MIAS database, contain variety mammograms. compare our model similar algorithms several state-of-the-art methods through comprehensive set experiments. approach, focus providing hybrid for accurate extraction mammography, physician can predict both potential disease stage type tumor.