ANFIS based Automated System for the Detection of Gadolinium Lesions in Brain MRI

Multiple sclerosis (MS) is a chronic disease that affects different parts of the central nervous system at different points of time. The central nervous system is made up of the brain and spinal cord. MS is an auto-immune disorder of the central nervous system characterized by the presence of Gadolinium lesions. In this paper, an intelligent technique is proposed for the automatic segmentation of Gadolinium lesions from the brain Magnetic Resonance Imaging (MRI) of MS patients. Since, manual interpretation of the lesions based on visual examination by radiologist/physician may lead to erroneous diagnosis when a large number of MRIs are analyzed; we propose an automated intelligent classification system. The proposed technique uses a trained classifier, Adaptive Neuro Fuzzy Interference System (ANFIS) to discriminate between the regions of MS lesions and non-MS lesion regions mainly based on the textural features. The proposed method consists of three stages, namely, preprocessing, feature extraction and classification. The classifier’s goal is to classify subjects as normal and abnormal brain MRI. The main contribution of the proposed technique described in this paper is the use of textural features like Gray level cooccurrence features, Local binary pattern features, Gray level based features, Histogram and Wavelet features to detect Gadolinium lesions in a fully automated approach that does not rely on manually delineating the lesions. The obtained results exhibit a sensitivity of 72% and maximum accuracy of 99% which assures that the proposed method would be viable for use in clinical practice for the detection of Gadolinium lesions in brain MRI of MS affected patients.