MACHINE LEARNING BASED CLINICAL DECISION SUPPORT SYSTEM FOR DIAGNOSIS OF NEURODEGENERATIVE DISEASES (MCADS-ND)

Abstract

Neuroimaging research is an interdisciplinary field that is impacting neuroscience, psychiatry, neurology, and other related disciplines. Despite the recent advancements in neuroimaging, medical diagnosis is still mainly based on a manual assessment of patient’s history and valuation of his/her observable signs and symptoms. Thus, the onus of confirmative diagnosis heavily relies only on the expertise of a physician. We have developed a battery of machine learning based systems to aid clinicians in differential disease diagnosis. The applied methodologies successfully achieved high prediction accuracy especially on images obtained from early-stages of the disease. In clinical settings, the induction of these algorithms as a decision support system could make a significant impact on treatment strategies especially by aiding early disease diagnosis.