Table 2.
Network | A | P | TPR | F 1 | AUC |
---|---|---|---|---|---|
3D AlexNet | 0.769 | 0.773 | 0.810 | 0.791 | 0.899 |
3D VGG | 0.769 | 0.800 | 0.762 | 0.780 | 0.860 |
3D ResNet | 0.872 | 0.808 | 1.000 | 0.894 | 0.931 |
Experiment 1: Quantitative comparison of the classification quality of three different networks for one-fold. The best result for each metric is marked bold. A = accuracy; P = precision; TPR = true positive rate; F1 = F1 score; AUC = area under the curve.