On the Detection of Alzheimer’s Disease and Mild Cognitive Impairment using Priority Feature Selection based Decision Tree Classifier

Alzheimer’s disease (AD) is one of the most common neurodegenerative dementias affecting the elderly population. Early detection of AD is crucial for improving quality of life, which has led to significant research interest worldwide. This study introduces an innovative approach to categorize brain MRI scans into two groups: mild cognitive impairment (MCI) and AD. The classification leverages volumetric information from four key features: cerebrospinal fluid (CSF), amygdala, caudate, and grey matter (GM). Using an information gain-based feature selection approach and a decision tree classifier, the proposed method is evaluated on ADNI dataset of brain MRI images. The method’s superiority is demonstrated by performance metrics such as accuracy, sensitivity, specificity, and F1-score, which show average gains of 12.02%, 8.52%, and 17.72% in accuracy, sensitivity, and specificity, respectively, when compared to state-of-the-art techniques.