Table 3.
Evaluation of proposed CNN and the fine-tuning trained models with preprocessing using proposed image enhancement.
| Method | Image dataset | Accuracy % | Sensitivity % | Specificity % | Precision % | F1-Score % |
|---|---|---|---|---|---|---|
| Proposed customized CNN | X-Ray | 98.60 | 98.40 | 98.50 | 98.30 | 98.35 |
| CT | 98.80 | 98.50 | 98.40 | 98.60 | 98.55 | |
| Pre-trained VGG16Net | X-Ray | 98.20 | 98.30 | 98.30 | 98.15 | 98.22 |
| CT scans | 98.10 | 98.10 | 98.20 | 98.00 | 98.05 | |
| Pre-trained AlexNet | X-Ray | 98.20 | 98.20 | 98.30 | 98.10 | 98.15 |
| CT scans | 98.40 | 98.10 | 98.20 | 98.00 | 98.05 |