Table 4.

Classification results of experiment 2: Using SMOTE balancing technique.

Dataset	CNN	ACC (%)	SEN (%)	SPE (%)	PPV (%)	NPV (%)	F1 (%)	AUC (%)
Dataset 1	DenseNet201	98.64	98.36	99.35	98.64	98.64	98.63	98.97
	CheXNet	97.51	97.50	98.75	97.51	97.51	97.50	98.20
	Xception	96.15	96.15	98.07	96.15	96.15	96.14	96.94
	ResNet152	95.46	95.64	97.73	95.46	95.46	95.46	96.60
	VGG19	98.19	98.19	99.09	98.19	98.19	98.19	98.63
	MobileNetV2	97.51	97.51	98.75	97.54	97.54	97.50	98.13
Dataset 2	DenseNet201	92.28	92.20	96.10	92.20	92.20	92.20	94.38
	CheXNet	91.19	91.19	95.59	91.18	91.18	91.12	93.42
	Xception	89.44	89.43	94.71	89.43	89.43	89.43	91.81
	ResNet152	91.82	91.82	95.91	91.82	91.82	91.82	93.84
	VGG19	92.08	92.07	96.03	92.08	92.08	92.07	93.95
	MobileNetV2	83.31	89.30	94.65	89.41	89.41	89.29	92.16
Dataset 3	DenseNet201	92.71	92.71	96.36	92.77	92.77	92.71	90.66
	CheXNet	92.02	92.02	92.02	96.01	96.01	92.10	89.54
	Xception	90.00	90.01	95.00	90.00	90.00	90.02	86.53
	ResNet152	91.32	91.32	95.66	91.32	91.32	91.32	88.93
	VGG19	93.66	93.66	96.83	93.66	93.66	93.66	92.00
	MobileNetV2	91.38	91.38	95.69	91.65	91.65	91.35	89.00