Table 1.
Comparative analysis of existing explainable artificial intelligence (XAI) and classification models for lung diseases.
| Methodology | Dataset | Explainability Models | Accuracy % |
|---|---|---|---|
| ResNet 101 [18] | 897 CT Scans | CAM, LIME | 85% |
| U-Net CNN [24] | 1478 X-rays | Grad-CAM, LIME | 83% |
| VGG16, ResNet [27] | 3975 CXRs | GSInquire | 83% |
| Xception [28] | 2235 chest X-rays | SHAP, Grad-CAM++ | 87% |
| DenseNet, ResNet, VGGNet [29] | 5959 CXRs | Grad-CAM++, LRP | 90% |
| DenseNet169 [30] | 787 CT Scans | Not Used | 85% |
| Proposed Mode | 787 CT Scans, 10,000 CXRs Scans | LIME | 93%, 97% |