In this study, a compartmental sparse feature selection method was used with feature parameter identified, and compared with classical voxel-based morphometry method (VBM) for classification of Alzheimer ’s disease (AD) from the healthy subjects. Our method had high classification accuracy for AD diagnosis and a strong linear correlation between the extracted feature parameter and volume of hippocampus obtained by VBM. The feature parameter of hippocampus had a higher linear correlation with mini-mental state examination (MMSE) score than volume of hippocampus with MMSE. Hence, compartmental sparse feature selection is an effective computer-aided diagnosis method to help clinician identify AD.
This abstract and the presentation materials are available to members only; a login is required.