THE SIGNIFICANCE OF ARTIFICIAL NEURAL NETWORKS ALGORITHMS CLASSIFICATION IN THE MULTIPLE SCLEROSIS AND ITS SUBGROUPS

Abstract: The study includes three subgroups of Multiple Sclerosis, namely RRMS, SPMS and PPMS as well as healthy individuals. The purpose of this study is to diagnose Multiple Sclerosis subgroups through Magnetic Resonance Imaging and Expanded Status Disability Scale. MRI and EDSS data belong to 139 volunteers, 120 of whom are MS patients (76 RRMS, 38 SPMS, 6 PPMS patients) and the remaining are healthy people. All subjects are between the ages of 20 and 55. Disability levels of MS symptoms are determined using Expanded Disability Status Scale. We have focused on three regions in the brain: brain stem, periventricular corpus callosum, and upper cervical regions. EDSS scores and number of lesions in these three regions are considered as the parameters of the ANN algorithms to determine the subgroups of the disease. The empirical results are examined taking two aspects into consideration. One of them is MRI and the other is EDSS data. These two elements are applied onto the input of Artificial Neural Networks Algorithms that are Feed Forward Back Propagation, Learning Vector Quantization and Radial Basis Function. The significance of these variables for the diagnosis of MS subgroups has been revealed as a result of this study in which algorithm has been utilized.

Keywords: Multiple Sclerosis, Magnetic Resonance Imaging, Expanded Disability Status Scale, Feed Forward Back Propagation, Learning Vector Quantization, Radial Basis Function.