Table 9. Performance ratios of the Ensemble of the full training and fine tuned of each CNN.
Architecture | AUC | Sensitivity (%) | Specificity (%) | B-Accuracy |
---|---|---|---|---|
VGG19 | 0.9219 | 83.64 | 89.45 | 86.35 |
RESNET | 0.9305 | 83.87 | 85.71 | 84.79 |
GOOGLENET | 0.9277 | 88.71 | 82.85 | 85.78 |
DENET DISC | 0.9083 | 81.45 | 88.35 | 84.90 |