Multiple Sclerosis Disorder Detection Through Faster Region-Based Convolutional Neural Networks

Multiple sclerosis is a leading brain disorder that highly affects the normal functions of the human body. Due to this disorder, protective coverings of neuron cells are get damaged, which causes disrupting the information flow inside the brain and other body parts. The early detection of multiple sclerosis helps healthcare practitioners to suggest a suitable treatment for the disease. The detection of multiple sclerosis is a challenging task. Many types of approaches had been proposed by the researchers and academicians for accurately detecting the brain lesions. Precisely, detecting the brain lesions is still a big challenge. Due to the recent innovations in the field of image processing and computer vision, healthcare practitioners are using advanced disease diagnosis systems for the prediction of disorders/diseases. Magnetic resonance imaging approach is used for the detection of various brain lesions by the neurosurgeons and neurophysicians. The computer vision approaches are playing a major role in the automatic detection of various disorders. In this research paper, the faster region-based convolutional neural networks approach is proposed based on computer vision and deep learning, using transfer learning for the detection of multiple sclerosis as a brain disorder. The proposed approach is detecting the damaged area inside the brain with higher precision and accuracy. The proposed model detects the multiple sclerosis brain lesions with 99.9% accuracy. Three DAGNetworks are used for training; there are Alexnet, Resnet18, and Resnet50. As compare to Alexnet and Resnet18, deep networks, the Resnet50 Pre-trained network performed well with higher accuracy of detection.