CLASSIFICATION OF NEURODEGENERATIVE DISEASES FROM EXTRACTION OF SALIENT BRAIN PATTERNS

Abstract

Neurodegenerative diseases causes a wide variety of mental symptoms whose evolution is not directly related to the analysis made by radiologists on basis of images, who can hardly quantify systematic differences. This paper presents a new automatic (Based on software program) image analysis method that reveals different brain patterns associated to the presence of neurodegenerative diseases, finding systematic differences and therefore grading objectively any neurological disorder. An accurate solution can be provided by using Alzheimer’s diseases based on saliency map characterization is carried out on database images. This paper gives automatic image analysis method and attempts an approach for classification of brain images to search for pathology and normality part of brain by extracting salient features of input brain image and the region of interest is identified using kernel k-means algorithm. A support vector machine (SVM) a supervised learning process is used for classification of AD, which is recognized on basis of blue color is normal brain part and red color is pathology related.