Table 4.

Performance analysis with pre-trained networks

Method	Accuracy	Sensitivity	Specificity	Recall	Precision	F1 score
ResNet50	91.60	96.67	89.07	81.56	96.67	88.47
GoogleNet	89.88	93.33	88.15	79.75	93.33	86.01
InceptionNet	92.59	96.67	90.56	83.56	96.67	89.69
SqueezeNet	91.48	97.78	88.33	80.73	97.78	88.44
AlexNet	90.74	97.41	87.41	79.46	97.41	87.52
Proposed algorithm	94.07	98.15	91.48	85.21	98.15	91.22