Automatic Computer-Aided Detection of Multiple Sclerosis (MS) Lesions on MR Images

Abstract

Multiple Sclerosis (MS) is a central nervous system (CNS) disorders resulting from damage to the myelin sheath which helps to ensure the transmission of messages between the brain and spinal cord. MS lesions occur in patients with damage to the myelin sheath. Progression of MS lesions is important for examining the disease. MS lesions often Magnetic Resonance Imaging (MRI) is determined and planned the follow-up / treatment processes. In this study, a computer-aided detection system has been proposed to diagnose MS lesions on FLAIR MR images. The proposed system uses Fuzzy-C Means (FCM) and morphological operations for segmentation of MS lesions. In the study, the lesions detected by aid of physicians and the lesions detected by means of the proposed system have been compared according to the Jaccard index. Similarity rate in the comparison operation on a total of 90 MR images is calculated as 91.2%. Consequently, it has been shown that the proposed system successfully detected the MS lesions.