Table 9. Comparison of the three-class classification studies from previous CXR-based COVID-19 studies.

Work	Number of Cases	Preprocessing	Approach	Performance (%)
Wang et al. [49]	266 COVID-19 8066 HC 5538 Pneumonia	DA	COVID-Net	Accuracy = 93.3 Sensitivity = 91 PPV = 98.9
Ucar et al. [43]	76 COVID-19 1538 HC 4290 Pneumonia	DA RGB format Normalizing	COVIDiagnosis-Net	Accuracy = 98.3 Specificity = 99.13 F1-Score = 98.3
Ozturk et al. [50]	127 COVID-19 500 HC 500 Pneumonia	N/A	DarkCovidNet (CNN)	Accuracy = 87.02 Specificity = 92.18 Sensitivity = 95.35 Precision = 89.96 F1-Score = 87.37
Li et al. [21]	179 COVID-19 179 HC 179 Pneumonia	Create a Noisy Snapshot Dataset	KTD framework (DenseNet121, ShuffleNetV2, MobileNetV2)	Accuracy = 84.3 AUCROC = 94
Punn et al. [44]	108 COVID-19 453 HC 515 Pneumonia	Class Balancing Methods Binary Thresholding Adaptive Total Variation Method	NASNetLarge	Accuracy = 98 Specificity = 95 Precision = 88 F1-Score = 89
Elasnaoui et al. [22]	6087 images (2780 Bacterial Pneumonia, 1724 Coronavirus (1493 Viral Pneumonia, 231 COVID-19)) 1583 HC	Intensity Normalization CLAHE Method DA Resizing	Inception ResNetV2	Accuracy = 92.18 Specificity = 96.06 Sensitivity = 92.11 Precision = 92.38 F1-Score = 92.07
Khobahi et al. [23]	99 COVID-19 8851 HC 9579 Pneumonia	DA	CoroNet (TFEN + CIN modules)	Accuracy = 93.50 Sensitivity = 90 Precision = 93.63 F1-Score = 93.51
Chowdhury et al. [74]	219 COVID-19 1341 HC 1345 Pneumonia	DA	PDCOVIDNet (CNN)	Accuracy = 96.54 Precision = 96.58 Recall = 96.59 F1-Score = 96.58
Chowdhury et al. [75]	589 COVID-19 8851 HC 6053 Pneumonia	DA	ECOVNet (pre-trained EfficientNet)	Accuracy = 94.68 Precision = 94.76 Recall = 94.68 F1-Score = 94.70
Perumal et al. [76]	183 COVID-19 8066 HC 5538 Pneumonia	N/A	INASNET (Inception Nasnet)	Accuracy = 94.3 Precision = 94.0 Recall = 94.0 F1-Score = 94.0
Proposed method	1093 COVID-19 1341 HC 1345 Pneumonia	Cropping, DA Histogram Equalization Constant Threshold Contouring	Feature-based Ensemble	Accuracy = 94.1 Precision = 94.5 Recall = 94.1 F1-Score = 94.0