TABLE III.
Results of Binary Disease Classification and Clinical Score Regression, With Models Trained on ADNI-1 and Tested on MIRIAD
| Binary Disease Classification |
Clinical Score Regression |
|||||
|---|---|---|---|---|---|---|
| Method | AD vs. NC |
MMSE |
||||
| ACC | SEN | SPE | AUC | CC | RMSE | |
| VBM | 0.884 | 0.913 | 0.826 | 0.921 | 0.623 | 5.271 |
| ROI | 0.870 | 0.913 | 0.826 | 0.918 | 0.699 | 5.369 |
| DSML-1 | 0.916 | 0.954 | 0.837 | 0.932 | 0.708 | 4.871 |
| DSML | 0.918 | 0.965 | 0.854 | 0.959 | 0.722 | 4.234 |
| DM2L-1 | 0.920 | 0.963 | 0.898 | 0.969 | 0.683 | 4.295 |
| DM2L | 0.937 | 0.946 | 0.932 | 0.986 | 0.736 | 4.136 |