Table 3. The classification performance of CNN models used in this study.

Models	Accuracy	Error	Recall	Specificity	Precision	F1-score	MCC
RestNet18	0.9568	0.0432	0.9675	0.9951	0.9382	0.9511	0.9471
GoogLeNet	0.9216	0.0784	0.9348	0.9912	0.8929	0.9061	0.9016
VGG19	0.8676	0.1324	0.9014	0.9851	0.8160	0.8361	0.8331
Inceptionv3	0.9572	0.0428	0.9672	0.9952	0.9234	0.9420	0.9389
MobileNetv2	0.9514	0.0486	0.9630	0.9945	0.9160	0.9356	0.9321
DenseNet201	0.9698	0.0302	0.9770	0.9966	0.9483	0.9607	0.9583
InceptionResNetv2	0.9734	0.0266	0.9734	0.9970	0.9606	0.9667	0.9638
EfficientNetb0	0.9644	0.0356	0.9720	0.9960	0.9261	0.9456	0.9433
ShuffleNet	0.9356	0.0644	0.9481	0.9927	0.8944	0.9161	0.9114
Modified ShuffleNet	0.9604	0.0308	0.9701	0.9906	0.9413	0.9527	0.9503