Table 6.

Comparative analysis of the proposed model with recently proposed models.

Dataset	Author	Model	ACC (%)	SEN (%)	SPE (%)	PPV (%)	F1 (%)	AUC (%)
Dataset 1	Chowdhury et al. [47]	DenseNet201	97.9	97.9	98.8	97.95	–	–
	Bassi et al. [49]	DenseNet121	98.3	–	–	98.3	98.30	–
	Proposed models	WCL$+$CheXNet	98.87	98.86	99.43	100	98.21	99.15
		SMOTE$+$DenseNet201	98.64	98.36	99.35	98.64	98.63	98.97
Dataset 2	Haghanifar et al. [79]	CheXNet	–	–	–	–	85	–
	Ozturk et al. [83]	DarkNet	87.02	–	–	–	–	–
	Proposed models	WCL$+$DenseNet201	93.58	93.58	96.79	93.58	93.58	95.38
		SMOTE$+$DenseNet201	92.28	92.20	96.10	92.20	92.20	94.38
Dataset 3	Wang et al. [44]	COVID-Net	92.4	88	–	91	–	–
	Oh et al. [30]	ResNet-18	91.9	–	–	76.9	–	–
	Proposed models	WCL$+$VGG19	94.87	94.87	97.43	94.92	94.87	94.86
		SMOTE$+$VGG19	93.66	93.66	96.83	93.66	93.66	92.00