Table 2 Comparative tables of our approach. The performance of our approach is compared to other methods on Dataset A: MODMA, Dataset B: EEG Psychiatric dataset, and Dataset C: EEG and Psychological Assesment dataset.
Dataset | Model | Accuracy (%) | Precision (%) | Recall (%) | F1-score | Number of channels |
|---|---|---|---|---|---|---|
Dataset A | MCT (our) | 89.84% | 87.36% | 89.41% | 88.37% | 5 |
NC+KNN24 | 83.96% | – | – | – | 13 | |
CNN+GRU25 | 89.63% | – | – | – | 19 | |
Dataset B | MCT (our) | 92.28% | 90.52 | 90.73% | 90.62% | 5 |
EN (baseline)21 | 87.59% | – | – | – | 19 | |
LSTM23 | 90% | 89.8% | 90% | 89.89% | 20 | |
Dataset C | MCT (our) | 87.40% | 86.17% | 87.22% | 86.69% | 5 |
Vit26 | 86.83% | 85.82% | 85.93% | 85.87% | 5 | |
LSTM27 | 74.62% | 71.83% | 72.37% | 72.09% | 5 | |
ResNet5028 | 86.94% | 86.37% | 84.49% | 85.41% | 5 |
