Table 2. Performance results of the models.

	VGG-16	ResNet-101	Inceptionv3	MobileNetV2
	%	%	%	%
Accuracy	86.6	82.8	78.7	77.6
Sensitivity	82.5	77.2	73.7	64.9
Specificity	89.6	87.0	81.8	87.0
Precision	85.5	81.5	75.0	78.7
F1 score	83.9	79.2	74.3	71.1
Area under the curve	0.92	0.88	0.84	0.80