Multiple Sclerosis Diagnosis Methods Using Machine Learning and Imaging Techniques

Multiple Sclerosis (MS) disease is immune disorder that destroys myelin in the nervous system and causes many complications including motor and sensory disorders. Nowadays, medical images including Magnetic Resonance Imaging (MRI) and Optical Coherence Tomography (OCT) are recognized as the basic tools in the diagnosis of MS disease. Due to the large amount of image data in this method, the use of machine learning methods, especially Neural Networks (NNs) plays an important role in image processing. This paper presents a comprehensive overview of different methods, which utilize NNs to MS diagnosis. This review presents the classical of NNs and Convolutional NNs (CNNs), which are used in the MS diagnosis. In addition, challenges, and recent developments in this field are presented, which provides directions for future researches in this field.

Keywords:

Multiple sclerosis disease , Magnetic resonance imaging , Machine learning , Neural networks , Deep learning