A hybrid neuro-fuzzy approach for brain abnormality detection using GLCM based feature extraction

Abstract

Brain tumor detection is an important task in medical field because it provides anatomical information of abnormal tissues in brain which helps the doctors in treatment planning and patient follow-up. In this paper an approach for detection and specification of anomalies present in brain images is proposed. The idea is to combine two metaphors: Neural Network and Fuzzy Logic. These two metaphors are combined in one system called Hybrid Neuro-Fuzzy system. This system enjoys the benefits of both Artificial Neural network system and Fuzzy Logic system and eliminates their limitations. The Neuro-Fuzzy system combines the learning power of Artificial Neural Network system and explicit knowledge representation of fuzzy inference system. The proposed system consists of four stages: data collection through various repository sites or hospitals, Pre processing of various brain images, Feature extraction using Gray Level Co-occurrence Matrix (GLCM) and classification of brain images through Hybrid Neuro-Fuzzy System. Experimental results illustrates promising results in terms of classification accuracy, specificity and sensitivity.