Multiple Sclerosis Prediction and Assessment Approaches: A Comparison Framework

Multiple Sclerosis (MS) is a complex neurological disorder of the Central Nervous System. It results in progressive loss of neurological functions and produces disability for the patient. This causes also a high cost for both the patient and the society. The etiology of MS is unknown and it is hard to predict the progression of the disease. There has been a number of machine learning based approaches proposed to address this problem. However, there is no framework that could be used to guide researchers to evolve new methods and strategies to better understand this disease. This motivated us to research previous works that attempt to predict diseases in general and MS and its progression in particular using machine learning approaches. Our goal is to compare current approaches and identify gaps present in this field of study. To allow a systematic comparison of current approaches, we developed an evaluation framework composed of a set attributes. These attributes were identified based on a thorough analysis of existing approaches and in consultation with prominent neurologists as well. The paper discusses a number of representative machine learning disease approaches against the framework. Analysis of the discussion highlights some open issues for future research.