Adaptive neuro-fuzzy inference system for classification of mammographic image using electromagnetism-like optimisation

Abstract

Breast cancer diagnosis system plays a vital role in medical field. This system helps the doctors to diagnose in much more efficient way. Breast cancer is a very common class of cancer among women. This paper mainly focuses on breast cancer recurrence problem, hybridising two methodologies; Electro-Magnetism-like Optimisation (EMO) and Adaptive Neuro-Fuzzy Inference System (ANFIS), to develop a good diagnosis system. EMO has been used as a multilevel segmentation algorithm which can effectively identify the threshold values of a digital image within the reduced number of iterations and decreasing the computational complexity. Original proposals show better results in diagnosing cancer affected cells to find the best features, whilst ANFIS algorithm is used as a classifier. ANFIS model combines the neural network adaptive capabilities and the fuzzy logic qualitative approach. The robustness of the proposed hybrid methodology is examined using classification accuracy, sensitivity, and specificity.