Table 7.

Performance of Architectures for Covid-19 detection by different metrics on all datasets containing X-ray and CT images.

Model	Accuracy	Precision	Recall	F1-Score	Specificity	AUC (%)
DenseNet	93.05	0.94	0.93	0.94	90.41	95.11
AlexNet	90.98	0.92	0.91	0.91	86.91	93.66
ResNet	92.02	0.93	0.93	0.93	90.30	94.70
CspNet [98]	90.71	0.92	0.91	0.91	87.14	93.67
VGG16	89.65	0.91	0.90	0.90	84.90	92.88
VGG19	89.24	0.91	0.89	0.90	83.64	92.27
CovXNet [42]	92.45	0.94	0.91	0.93	88.85	93.86
CoroNet [40]	92.23	0.94	0.92	0.93	87.28	94.43
CovidXrayNet [41]	95.30	0.96	0.96	0.96	92.93	96.85
DarkCovidNet [39]	90.59	0.92	0.91	0.91	88.05	93.06
Proposed (No DDC)	92.19	0.93	0.92	0.93	87.69	94.40
Proposed+ DataAug.	91.92	0.94	0.91	0.90	89.90	93.10
Proposed (No GB)	93.36	0.94	0.94	0.94	91.73	95.50
Proposed(CovidDWNet+GB)	96.32	0.97	0.97	0.97	95.17	97.67