Table 6. Comparisons of the proposed method with state-of-the-art methods for PD diagnosis.
| Method | Subjects |
Methodology | Modalities | PD vs. NC (%) | |
|---|---|---|---|---|---|
| PD | NC | ||||
| Prashanth18 | 493 | 181 | Linear Classifier | SPECT | 92.3 |
| Prashanth18 | 493 | 181 | Non-linear Classifier | SPECT | 96.1 |
| Salvatore31 | 28 | 28 | PCA + SVM | MRI | 85.8 |
| Adeli-M2 | 56 | 56 | Robust Feature Sample LDA | MRI | 84.1 |
| Ours | 369 | 169 | Kernel-based Feature Selection and Max-Margin Classification | MRI + SPECT | 97.5 |
The table includes number of subjects and the methodologies used by different methods. Also, the neuroimaging modalities used by each study are provided.