UNLOCKING NEUROLOGICAL MYSTERIES: MACHINE LEARNING APPROACHES to EARLY DETECTION of ALZHEIMER'S DISEASE

Abstract

Dementia is a clinical illness that becomes more common as people get older. It is defined by a decline in cognitive abilities across several domains and eventually impacts everyday functioning. Consequently, this leads to a decline in autonomy, impairment, dependence on assistance, and ultimately, mortality. Alzheimer's disease (AD) is responsible for 50–80% of all occurrences of dementia, and its occurrence increases by a factor of five every five years beyond the age of 65. Given the availability of health data and the decrease in data processing costs, it is now feasible to detect Alzheimer's disease at an early stage. The objective of this study is to classify individuals as either Alzheimer's sufferers or healthy individuals by employing various machine learning techniques. The OASIS-2 dataset, which consists of longitudinal MRI data from both nondemented and demented older adults, was utilized for this study. Given its potential for early detection of Alzheimer's dementia, the study is anticipated to enhance clinical decision support systems pertaining to modifiable risk factors.