Table 9.
Comparison of CoVNet-19 with other state-of-the-art methods.
| Reference study | Model architecture | X-ray dataset | 2-Class accuracy | 3-Class accuracy |
|---|---|---|---|---|
| Ioannis D. et al. [7] | VGG19 | 224 COVID ＋, 504 Normal. 700 Pneumonia |
98.75% | 93.48% |
| Wang et al. [8] | COVID-Net network | 183 COVID ＋, 8066 Normal, 5538 COVID - |
N/A | 93.33% |
| Asif Iqbal Khan et al [9] | CoroNet | 290 COVID ＋, 1203 Normal. 1590 Pneumonia |
99.00% | 95.00% |
| Narin et al. [10] | ResNet50 | 50 COVID ＋, 50 COVID - |
98.00% | N/A |
| Sethy et al. [11] | ResNet50 + SVM | 25 COVID ＋, 25 COVID - |
95.38% | N/A |
| Ozturk et al. [12] | DarkCovidNet | 127 COVID ＋, 500 Normal, 500 Pneumonia |
98.08% | 87.02% |
| Ezz El-Din Hemdan [16] | VGG19 | 25 COVID ＋, 25 COVID - |
90.00% | N/A |
| Mesut Tougaçar et al. [18] | MobileNetV2+ SqueezeNet + SVM | 295 COVID ＋, 65 Normal, 98 Pneumonia |
N/A | 99.27% |
| Ferhat Ucar et al. [19] | Deep Bayes SqueezeNet | 76 COVID ＋, 1583 Normal, 4290 Pneumonia |
N/A | 98.30% |
| Mahesh Gour [42] | Stacked CNN | 270 COVID ＋, 1139 Normal, 1355 Pneumonia |
N/A | 92.74% |
| Proposed Model | CoVNet-19 |
1628 COVID ＋ 2341 Normal, 2345 Pneumonia |
99.71% | 98.28% |
**N/A: Authors did not perform the specified classification.