In the last 10 years, detection of Alzheimer’s disease based on brain T1-weighted Magnetic Resonance Imaging (MRI) have been a highly sought goal in the neuroscientific community. However, the methods were assessed on different datasets and not always publicly available ones making reproducibility and validation impossible. Here, we evaluated five deformation-based features (Jacobian determinant and trace, modulated grey matter (GM), deformation vector norm and geodesic anisotropy) and three feature selection processes (Pearson correlation, Bhattacharyya distance and Welch’s t-test) using the 416 subjects (316 controls and 100 patients) from the OASIS database. Our objective was to correctly discriminate between controls (CN) and Alzheimer’s disease patients (AD). For this task we used a fully and non-fully independent stratified 10-fold cross-validation and linear SVM. Our best mean detection result was 79.43% of precision, 96.67% of sensitivity and 97.37 of area under the ROC curve.We also show discriminant voxel site locations found with each measure.
IOS Press, Inc.
6751 Tepper Drive
Clifton, VA 20124
Tel.: +1 703 830 6300
Fax: +1 703 830 2300 email@example.com
(Corporate matters and books only) IOS Press c/o Accucoms US, Inc.
For North America Sales and Customer Service
West Point Commons
Lansdale PA 19446
Tel.: +1 866 855 8967
Fax: +1 215 660 5042 firstname.lastname@example.org