MULTIPLE SCLEROSIS DETECTION USING IMAGE PROCESSING TECHNIQUES

Abstract

Cancer is the main and the most crucial reason behind the huge number of deaths across the world, and in the developing countries like India the inception of sickness equivalent to cancer will disgrace the economy of the country. The disease like cancer ,ore importantly the disease Multiple Sclerosis is the only and the crucial reason behind the huge number of death rates and it is autoimmune disease which cause the most mortality rates due to being non curable disease. Computer tomography scan like MRI (Magnetic Resonance Imaging) is employed by radiologist to discover lesions present in the gray or the white matter of the brain which causes neurological disorders and is used to artifact the growth of the symptoms. Seeable version of the information that is obtained which enable the one to detection of Multiple Sclerosis at subsequent phases, therefor resulting in the prolonged treatment of the disease that ultimately increases the number of deaths caused by the different types of cancers. Because of this reason the Image Processing Toolbox which has many different tools is employed in order to detect the Multiple Sclerosis in advance. In the presented work, multiple Sclerosis is detected in the brain MRI image and the detection algorithm for the same is accomplished using the brain MRI image which is segmented with the help of Fuzzy C- means Segmentation algorithm in addition with some of the morphological operations that are employed for the accurate segmentation of Brain Region of Interest, from which Gray-Level Co-occurrence Matrix Algorithm (GLCM) features are extracted using discrete wavelet transform (DWT) algorithm and then finally used for classification of Lesions by traditional machine learning Classifier which is support vector machine (SVM) along with the Deep learning classifier which Convolutional neural Network (CNN).