Table 7.
Classification performance using an individual CNN-based model (unit: %).
Fold Index | VGG-Based Method | Inception-Based Method | DenseNet-Based Method | ||||||
---|---|---|---|---|---|---|---|---|---|
Sens | Spec | Accuracy | Sens | Spec | Accuracy | Sens | Spec | Accuracy | |
First Fold | 14.565 | 97.010 | 67.370 | 22.065 | 85.723 | 62.837 | 27.065 | 95.729 | 71.043 |
Second Fold | 62.412 | 75.491 | 67.403 | 6.790 | 95.138 | 40.506 | 13.641 | 91.601 | 43.392 |
Average | 51.979 | 85.088 | 67.392 | 10.121 | 90.938 | 47.745 | 16.567 | 93.442 | 52.356 |