Automated segmentation of multiple sclerosis lesions using statistical approach

Abstract

This paper describes a statistical approach for segmenting multiple sclerosis lesions (tumors) from magnetic resonance imaging (MRI) images. Proposed method detects and segments the areas inside the brain that are affected by tumors. Tumor regions are the areas of higher intensity in comparison to normal tissue. Our automated method gives satisfactory results showing that the proposed method is capable of segmenting multiple sclerosis lesions of different shapes and intensities. In order to show the efficacy of proposed approach, experimental results are compared with the results of other algorithm and also with the results of manual segmentation performed by experts. Index Terms-Multiple sclerosis (MS) lesions, intra -class variance, MRI image, skull extraction, morphological operations.IEEE Sponsored 2nd International Conference on Innovations in Information Embedded and Communication Systems ICIIECS'15 978-1-4799-6818-3/15/$31.00

Review of advanced computational approaches on multiple sclerosis segmentation and classification

“…Moreover, the results were represented utilising the common execution parameters PSNR and DICE. The improved execution of the outcomes demonstrated the viability of the methodology [43].…”