Survey on Feature Detection of Alzheimer’s disease in Early Stage by Particle Size Distribution in Feature Selection Optimization Procedure

Abstract

The rising incidence of Alzheimer's Disease (AD) and moderate impairments of scan results in the modern world have piqued the interest of scientists in the field of neuroimaging-based diagnostics. Neuroimaging allows for the quantification of pathological alterations in the brain that has been associated to AD. Through the use of categorization frameworks, which provide diagnostic and prognostic tools, these measurements have been quickly included into the signatures of AD in recent years. The purpose of this article is to summarise research on Alzheimer's disease that used optimization strategies for feature selection. To address the problem of excessive model complexity when using ML techniques, this work presents a novel approach to feature selection. Several stages of Alzheimer's disease and a state of altered brain function that is clinically similar to AD but less severe have been described. The effectiveness of an AD's classification in these approaches is evaluated using a wrapper-based feature selection mechanism. Then, a suggested Social Spider Metaheuristic algorithm has been employed to zero down on the most crucial characteristics for making a correct AD diagnosis.