Table 5 Six evaluation metrics for 3 models
Model | %(95% CI) | |||||
|---|---|---|---|---|---|---|
Precision | Recall | Specificity | F1 Score | kappa | Accuracy | |
ResNet50 | 93.4(91.5–95.2) | 89.5(83.9–95.1) | 98(95.3–100) | 91.3(87.9–94.7) | 0.872 | 0.915 |
ResNet101 | 91.2(88.3–94.1) | 86.3(78.1–94.4) | 97.5(94.2–100) | 88.5(82.9–94) | 0.845 | 0.903 |
Fusion model | 94.5(92.8–96.3) | 89(82.1–95.8) | 98(95–100) | 91.4(87.5–95.4) | 0.878 | 0.9237 |