Table 2 Comparison of the performance of EEG and audio data fusion models at different depths of GCN layers
From: An adaptive multi-graph neural network with multimodal feature fusion learning for MDD detection
Modality | Method | ACC(\(\%\)) | PRE(\(\%\)) | REC(\(\%\)) | F1 Score(\(\%\)) |
|---|---|---|---|---|---|
Multimodal | MS2-GNN31 | 86.49 | 82.35 | 87.50 | 84.85 |
Ahmed et al.34 | 95.78 | 93.45 | 95.64 | 94.53 | |
Effnetv2s13 | 93.07 | 92.92 | 91.76 | 93.92 | |
Mobile-Net13 | 83.89 | 78.81 | 77.94 | 78.07 | |
Hu et al.35 | 80.59 | - | - | - | |
EMO-GCN | 96.76 | 96.26 | 95.37 | 95.81 | |
EEG | Tasci et al.36 | 83.96 | 86.76 | 76.14 | 81.10 |
SGP-SL37 | 84.91 | 80.77 | 87.50 | 84.00 | |
Soni et al.38 | 88.80 | 86.60 | 87.20 | 87.10 | |
Shen et al.39 | 72.25 | - | 81.88 | - | |
Sun et al.40 | 84.18 | - | 78.29 | - | |
EMO-GCN-\(\alpha\) | 90.06 | 90.20 | 88.46 | 89.32 | |
Audio | GNN-SDA41 | 82.70 | 82.60 | 79.20 | 80.90 |
Gheorghe et al.42 | 84.16 | 85.30 | 83.80 | 84.00 | |
Sun et al.28 | 90.35 | 88.25 | 90.33 | 89.15 | |
Chen et al.43 | 83.40 | 83.50 | 76.80 | 80.00 | |
Das et al.44 | 90.47 | 89.53 | 89.43 | 89.47 | |
EMO-GCN-\(\beta\) | 90.48 | 92.36 | 90.48 | 91.41 |
