Search

Multiple Sclerosis (MS), an autoimmune condition in which the immune cells of the body destroys the covering (myelin sheath) surrounding the nerves, hinders the brain from effectively communicating with the body. If progressed to the later stages, the condition can lead to irreparable deterioration of the nerves. Neurologists use Magnetic Resonance Images (MRI) to classify them. Due to the complexity of the brain and MRI scan images, manual examination might be time consuming. To aid doctors across the world and reduce the time taken to diagnose this disease, Convolutional Neural Networks (CNNs) have been suggested. Moreover, this paper is a comparative study using two CNN Architectures done on MRI scan images for MS diagnosis. The dataset constitutes MRI scan images of MS patients in 2 different orientations and healthy brain MRI images. Expeditious diagnosis leads to an earlier medicament which can help contain the spread of the disease. Metrics like accuracy and other evaluation criteria are considered for the comparison of the CNN models.