Table 5 Binary classification (melanoma vs. non-melanoma).
From: Advancing skin cancer diagnosis with deep learning and attention mechanisms
Metric | Proposed model | UNet model | DenseNet model | ResNet model |
|---|---|---|---|---|
Accuracy | 97.8 | 88.4 | 90.1 | 91.5 |
Precision | 98.1 | 85.2 | 89.5 | 90.3 |
Sensitivity | 98.3 | 85.5 | 88.2 | 89.1 |
Specificity | 97.5 | 91.2 | 92.6 | 93.0 |
F1-Score | 98.2 | 85.3 | 88.9 | 89.6 |
Dice Score | 98.8 | 85.0 | 87.0 | 89.0 |
IoU | 99.4 | 82.0 | 87.0 | 89.0 |
