Classification and Feature Extraction of Alzheimer’s from Structural MRI

Abstract: Alzheimer’s Disease (AD) is a neuro-degenerative disease which causes cell death and damage for the tissues in the brain. AD condition causes memory loss which disrupts their daily activity and they ultimately gets detached from their own surroundings. The Progression of Alzheimer’s disease can be controlled by early detection. The proposed methodology uses structural Magnetic Resonance Image (MRI) for AD detection. From MRI data AD can be classified into three different stages such as Cognitive Normal (CN), Mild Cognitive Impairment (MCI) and Alzheimer’s Disease. Initially MRI image is preprocessed which includes contrast enhancement and skull stripping. Then preprocessed image is segmented to gray matter, white matter and cerebrospinal fluid. After feature extraction, by some machine learning models the brain image is classified as normal vs Alzheimer’s.

Keywords: Alzheimer’s Disease, Magnetic Resonance Image, Support Vector Machine, Random Forest.