Table 11.

Comparison of evaluation performance of the proposed model against the state-of-the-art models that used the same dataset. This comparison is conducted using the multi-class prediction scenario with four respiratory class diseases: COVID-19 vs. Pneumonia vs. Lung Opacity vs. Normal.

Reference	AI Architecture	ACC	PRE	SEN	F1_Score
(Khan et al., 2020)	CoroNet	0.89	0.90	0.96	0.90
(Wang et al., 2020, Wang et al., 2020)	COVIDNet	0.90	0.91	0.91	0.91
(Lee et al., 2019),	Multiscale Attention Guided Network	0.92	0.93	0.92	0.92
(Shi et al., 2021)	Teacher Student Attention	0.91	0.92	0.91	0.91
(Mondal 2022)	Local Global Attention Network	0.96	0.96	0.96	0.96
(Khan et al., 2022a, Khan et al., 2022b) (Strategy 1)	EfficientNetB1	0.92	0.92	0.95	0.93
	NasNetMobile	0.89	0.89	0.92	0.91
	MobileNetV2	0.90	0.92	0.92	0.92
(Khan et al., 2022a, Khan et al., 2022b) (Strategy 2)	EfficientNetB1	0.96	0.97	0.97	0.98
	NasNetMobile	0.95	0.95	0.95	0.95
	MobileNetV2	0.94	0.94	0.95	0.95
Ours (2023)	The Proposed XAI Model	0.98	0.96	0.96	0.96