A Survey on Alzheimer’s Disease Prediction Using Deep Learning Algorithms

Abstract

Cognitive function, behaviour, and memory are all affected by Alzheimer’s disease, a form of dementia. The condition affects millions of adults in their mid-60 s each year. Alzheimer’s disease cannot be managed only with medication. It is still difficult to predict which patients with mild cognitive impairment (MCI) would later experience dementia brought on by Alzheimer’s disease (AD). It contains comprehensive details about Alzheimer’s disease and guidelines for making a medical diagnosis. Early AD detection has never been easy, but academics that look at connected computers are continuously trying to improve it. Finding a deep model that is more accurate and effective for diagnosing AD than conventional machine learning approaches is of great importance because deep learning has emerged as the method of choice for analysing medical images. In the first half of the survey, we carefully assess the potential predictive usefulness of DL techniques for Alzheimer’s disease. Second, compare and evaluate the most effective core DL methods for Alzheimer’s disease prediction. The best core prediction method’s parameter tuning analysis was performed to achieve improved accuracy using the statistical variation tool ANOVA. For this review, several field studies were retrieved using various research sources. As the study concludes, it discusses upcoming discoveries and difficulties in identifying and categorizing Alzheimer's disease.