Abstract
Alzheimer’s Disease (AD) is a progressive neurodegenerative disorder and a significant global health problem. Standard diagnostic tests frequently miss the disease until after substantial cognitive impairment, thus minimizing the time for optimal intervention and management. In light of this problem, the current research suggests a low-cost machine learning–based solution for early AD prediction. In contrast to costly and time-consuming imaging technologies, in the present research, a structured tabular dataset is used, consisting of clinical histories, cognitive test scores, and other non-imaging factors. Comparative analysis revealed that the Extreme Gradient Boosting (XGBoost) model exhibited optimal performance with maximum accuracy for predicting the disease. These results indicate huge potential for machine learning in the development of scalable, low-cost, and reliable diagnostic systems, thus enabling early risk identification and better patient results in the war on Alzheimer’s disease.
View Research Publication: