An Experimental Study on Evaluating Alzheimer's Disease Features using Data Mining Techniques

Abstract

Alzheimer’s disease (AD) predominantly affects the elderly population with symptoms including, but not limited to, cognitive impairment and memory loss. Predicting AD and mild cognitive impairment (MCI) can lengthen the lifespan of patients and help them to access necessary medical resources. One potential approach to achieve an early diagnosis of AD is to use data mining techniques which explore various characteristic traits related to MCI, cognitively normal (CN), and AD subjects to build classifiers that reveal important contributors to the disease. These classifiers are used by physicians during the AD diagnostic process in a clinical evaluation. In this research, we compare between different data mining algorithms through empirical data approach to deal with the AD diagnosis. Experimental evaluation, using attribute selection methods, and classifiers from rule induction and other classification techniques have been conducted on data from the Alzheimer’s Disease Neuroimaging Initiative (ADNI-MERGE). The results illustrate the good classification performance of classifiers with rules in predicting AD.