A Novel Computer Vision Classifier Identifying Alzheimer's Stages from MRI Scans

Alzheimer's disease is a fatal neurodegenerative disease causing neural destruction whose symptoms peak at old age and worsen with time. Currently, over 55 million people worldwide suffer from Alzheimer's and dementia, 60% of whom reside in low-income countries. Machine learning technology diagnosing Alzheimer's disease has the potential to allow earlier diagnosis for treatment and prevention, and allows Alzheimer's treatment to reach low income communities. Our study explores a machine learning and deep learning classification algorithm to identify Alzheimer's based on magnetic resonance imaging brain scans. Using a balanced dataset, our classification algorithm was able to use MRI scans to sort out patients with no impairment, very mild impairment, mild impairment, and moderate impairment with an accuracy of 96%. Unlike other contemporary machine learning based classification for Alzheimer's, our study uses the more definitive method of MRI scans as opposed to patient statistics, as well as detects different levels of disease progression, allowing for stage-based treatment.