Table 13 Summary of PD detection studies that used the same publicly available PD datasets.
References | FE methods | Classifier(s) | Dataset | Classification | Accuracy (%) |
|---|---|---|---|---|---|
– | CNN + LSTM | UNM | Off-PD versus On-PD | 99.2 | |
Gabor transformation + 2D-CNN | SanDiego | Off-PD vs. HC On-PD vs. HC Off-PD vs. on-PD | 99.44 98.84 92.60 | ||
Smoothed pseudo-Wigner Ville distribution + CNN | SanDiego | Off-PD vs. HC On-PD vs. HC | 99.84 100.0 | ||
PSD | Hyperplanes | UNM | Off-PD versus HC | 85.40 | |
WT + statistical measures | SVM | SanDiego | Off-PD versus HC On-PD versus HC | 96.13 97.65 | |
CSP + LogEn | KNN | UNM (close/open) | Off-PD versus HC On-PD versus HC Off-PD versus on-PD | 98.81/99.01 98.77/98.85 98.73/98.97 | |
SVM KNN KNN | SanDiego | Off-PD versus HC On-PD versus HC Off-PD versus on-PD | 99.41 95.76 97.52 | ||
Present study | DWT + TShEn DWT + ThEn DWT + TShEn | SVM KNN SVM | UNM (close/open) | Off-PD versus HC On-PD versus HC Off-PD versus on-PD | 99.51/98.62 99.52/99.20 99.39/98.79 |
DWT + TShEn DWT + TShEn DWT + ThEn DWT + SuEn DWT + TShEn | KNN | SanDiego | On-PD versus HC Off-PD versus on-PD Off-PD versus HC Off-PD versus HC Off-PD versus HC | 94.21 98.84 99.72 99.66 99.89 | |
