Table 4 Comprehensive evaluation of the proposed model against existing frameworks for COVID-19 diagnosis across datasets 1, 2, 3, and 4.
From: Blockchain enabled collective and combined deep learning framework for COVID19 diagnosis
Dataset | Models | Precision | Recall | Specificity | F1-Score | Accuracy |
---|---|---|---|---|---|---|
Dataset 1 | EfficientNetB425 | 0.9058 | 0.9258 | 0.9014 | 0.9157 | 0.9254 |
Deep CNN31 | 0.8954 | 0.9152 | 0.8723 | 0.9052 | 0.9158 | |
Blockchain-basedfederated learning38 | 0.9258 | 0.9261 | 0.9154 | 0.9259 | 0.9174 | |
XGBoost26 | 0.9158 | 0.9014 | 0.8274 | 0.9085 | 0.9126 | |
CAP-CNN41 | 0.9314 | 0.9409 | 0.9026 | 0.9361 | 0.9417 | |
LSTM+blockchain14 | 0.9587 | 0.9414 | 0.8858 | 0.9500 | 0.9514 | |
Bi-LSTM+blockchain33 | 0.9458 | 0.9325 | 0.9087 | 0.9391 | 0.9418 | |
CLCD-Block (Proposed) | 0.9879 | 0.9884 | 0.9787 | 0.9881 | 0.9871 | |
Dataset 2 | EfficientNetB425 | 0.8852 | 0.8987 | 0.6547 | 0.8919 | 0.9021 |
Deep CNN31 | 0.8987 | 0.9185 | 0.8452 | 0.9085 | 0.8971 | |
Blockchain-basedfederated learning38 | 0.9189 | 0.9085 | 0.4521 | 0.9137 | 0.9154 | |
XGBoost26 | 0.8947 | 0.8547 | 0.5214 | 0.8742 | 0.9142 | |
CAP-CNN41 | 0.8754 | 0.8689 | 0.7524 | 0.8721 | 0.8725 | |
LSTM+blockchain14 | 0.9125 | 0.9142 | 0.9014 | 0.9133 | 0.9147 | |
Bi-LSTM+blockchain33 | 0.9248 | 0.9187 | 0.8549 | 0.9217 | 0.9097 | |
CLCD-Block (Proposed) | 0.9768 | 0.9776 | 0.9658 | 0.9772 | 0.9729 | |
Dataset 3 | EfficientNetB425 | 0.9154 | 0.9024 | 0.8874 | 0.9089 | 0.9285 |
Deep CNN31 | 0.8989 | 0.8958 | 0.8254 | 0.8973 | 0.9105 | |
Blockchain-basedfederated learning38 | 0.9198 | 0.9052 | 0.4291 | 0.9124 | 0.9125 | |
XGBoost26 | 0.8245 | 0.8147 | 0.8258 | 0.8196 | 0.8198 | |
CAP-CNN41 | 0.8958 | 0.8578 | 0.82531 | 0.8764 | 0.8874 | |
LSTM+blockchain14 | 0.9147 | 0.9058 | 0.8685 | 0.9102 | 0.9025 | |
Bi-LSTM+blockchain33 | 0.9258 | 0.9158 | 0.8978 | 0.9208 | 0.9154 | |
CLCD-Block (Proposed) | 0.9731 | 0.9649 | 0.9821 | 0.9690 | 0.9835 | |
Dataset 4 | EfficientNetB425 | 0.9154 | 0.8849 | 0.7584 | 0.8999 | 0.9058 |
Deep CNN31 | 0.8799 | 0.8971 | 0.8542 | 0.8884 | 0.9254 | |
Blockchain-basedfederated learning38 | 0.9254 | 0.9230 | 0.8756 | 0.9242 | 0.9148 | |
XGBoost26 | 0.7587 | 0.7895 | 0.7258 | 0.7738 | 0.7689 | |
CAP-CNN41 | 0.8782 | 0.8687 | 0.8658 | 0.8734 | 0.8725 | |
LSTM+blockchain14 | 0.9158 | 0.9021 | 0.9098 | 0.9089 | 0.9029 | |
Bi-LSTM+blockchain33 | 0.9236 | 0.9317 | 0.9157 | 0.9276 | 0.9256 | |
CLCD-Block (Proposed) | 0.9792 | 0.9731 | 0.9879 | 0.9761 | 0.9792 |