Alzheimer's Disease Classification Using Deep Learning

Abstract

Alzheimer's disease is a neurological condition that primarily impacts older person's memory and is incurable. Worldwide, Alzheimer's disease primarily impacts adults over 65. This condition requires an early diagnosis in order to be accurately detected. Due to the high number of individuals who come with the illness, manual determination by medical professionals is risky and laborious. There is a require for superior precision in early conclusion approaches despite the fact that a variety of strategies have been used to diagnose and categorise Alzheimer's disease. Such representations are capable of being learned from data by deep learning techniques. In the proposed study, transfer learning with ResNet-50 and Fastai will be used to accomplish multilayer categorization of Alzheimer's illness, i.e., Mild Demendia, Moderate Demendia, Non Demendia, and Very Mild Demendia. This technique results in high projected accuracy, a major improvement over past studies and ample evidence of the effectiveness of the suggested strategies.