Table 8.
Comparison of DenseCapsNet with classical CNN methods.
| Methods | Number of cases | Accuracy |
|---|---|---|
| GoogLeNet Loey et al. [38] | 69 COVID-19 + 79 Normal + 79 BP + 79 VP | 80.56 |
| Ozturk et al. [39] | 125 COVID-19 + 500 Pneumonia+ 500 No finding | 87.02 |
| Mahmud et al. [40] | 305 COVID-19 + 305 VP+ 305 BP | 89.6 |
| Hemdan et al. [14] | 25 COVID-19 (+) 25 Normal | 90.0 |
| Apostolopoulos et al. [41] | 224 COVID-19 + 700 Pneumonia +504 Normal | 93.5 |
| Heidari et al. [42] | 415 COVID-19 + 5179 CAP + 2880 Normal | 94.0 |
| Proposed | 250 COVID-19 + 250 OP + 250 Normal | 90.7 |
| 250 COVID-19 + 250 Non COVID-19 | 98.5 |
Note: BP, OP and VP stand for bacterial pneumonia, other pneumonia and viral pneumonia respectively.