Early Detection of Mild Cognitive Impairment using Neuropsychological Data and Machine Learning Techniques

This is the authors' abstract. We don't add key points for this paper.

Individuals with mild cognitive impairment (MCI) have a much higher risk of developing Alzheimer's disease (AD) or other types of dementia. Early detection of MCI is critical to identify these high-risk candidates for proper management and early interventional treatment. In this paper, a neuropsychological dataset from the Alzheimer's Disease Neuroimaging Initiative (ADNI), one of the largest public AD database, was used. Four different machine learning algorithms were investigated to classify MCI from demographically matched cognitively normal (CN) older adults. In total, there were 620 subjects, including 391 MCI and 229 CN. A comprehensive neuropsychological battery was administered to all subjects, and 3 tests were selected for this study: Alzheimer's Disease Assessment Scale-Cognitive subscale (ADAS-Cog), Mini-Mental State Examination (MMSE), and Functional Activities Questionnaire (FAQ). Of all the classifiers tested, the Support Vector Machine (SVM) classifier with radial basis function (RBF) kernel had the best performance for MCI detection, with an average classification accuracy of 88.06% and the area under the receiver operating characteristic curve (AUC) of 0.945.