Table 5.
Comparison of previous works with the proposed work
| Authors | Number of cases | Method/s used | Accuracy |
|---|---|---|---|
| Hemdan et al. (2020) | 25 COVID-19(+) 25 Normal |
COVIDX-Net | 90.0% |
| Zhang et al. (2020) | 100 COVID-19(+) 1431 COVID-19(−) |
18-layer residual CNN | 96% |
| Ozturk et al. (2020) | (a) 125 COVID-19(+) 500 No Findings |
DarkCovidNet | (a) 98.08% |
| (b) 125 COVID-19(+) 500 Pneumonia 500 No Findings |
(b) 87.02% | ||
| Oh et al. (2020) | 8851 Normal 6012 Pneumonia 180 COVID-19(+) |
Patch-based CNN | 88.9% |
| Abbas et al. (2020) | 80 Normal 105 COVID-19 11 SARS |
DeTRAC | 95.12% |
| Proposed method | 795 COVID-19(+) 795 Normal 711 Others |
VGG16, DenseNet121, Xception, NASNet, EfficientNet | 79.01%, 89.96%, 88.03%, 85.03%, 93.48% |