Table 3.
NC classification results of machine learning methods and deep learning methods on AS-OCT image dataset (The best results are marked in bold)
| Method | ACC | PR | Sen | F1 |
|---|---|---|---|---|
| SVM [16] | 82.19 | 83.64 | 85.05 | 84.29 |
| NB | 80.62 | 81.28 | 84.17 | 82.22 |
| LR | 81.80 | 85.36 | 82.65 | 83.84 |
| DT | 80.62 | 83.03 | 83.01 | 82.98 |
| RF [16] | 82.60 | 84.49 | 84.64 | 84.52 |
| Adaboost | 75.55 | 83.53 | 79.52 | 76.84 |
| XGboost | 82.18 | 84.45 | 84.18 | 84.31 |
| GraNet [15] | 85.05 | 85.66 | 87.25 | 86.37 |
| VGG19 | 84.38 | 86.41 | 85.70 | 85.94 |
| ResNet34 | 83.78 | 85.57 | 86.02 | 85.71 |
| ResNeXt29 | 84.76 | 87.6 | 85.92 | 86.63 |
| GoogleNet | 82.48 | 86.28 | 84.21 | 84.03 |
| EfficientNet | 84.42 | 86.12 | 86.30 | 86.11 |
| SKNet | 85.68 | 88.22 | 86.78 | 87.32 |
| SENet34 | 85.47 | 87.44 | 87.18 | 86.83 |
| BAM | 83.19 | 86.66 | 84.79 | 85.07 |
| GCA-Net-18 [17] | 83.66 | 85.97 | 84.68 | 85.26 |
| ECANet-18 | 85.09 | 86.45 | 86.32 | 86.38 |
| SPANet-18 | 85.22 | 88.13 | 86.37 | 86.92 |
| CBAM-ResNet18 | 84.54 | 87.20 | 85.88 | 86.39 |
| MPANet-18-C | 86.40 | 89.02 | 88.78 | 88.02 |
| MPANet-34-C | 86.61 | 88.07 | 89.02 | 88.31 |
| MPANet-18 | 86.70 | 88.26 | 89.06 | 88.59 |
| MPANet-34 | 86.99 | 88.42 | 89.09 | 88.70 |