Table 1.

Performance of ViTCNX and the other six models under the binary classification.

Metrics	Precision	Recall	Accuracy	F1-score	AUC	AUPR
EfficientNetV2	0.9920	0.3293	0.5875	0.4945	0.9609	0.9738
ConvNeXt	0.9650	0.9894	0.9715	0.9770	0.9952	0.9968
DenseNet	0.9788	0.9814	0.9756	0.9801	0.9973	0.9983
Swin Transformer	0.9587	0.9548	0.9471	0.9568	0.9911	0.9945
ResNet-50	0.9892	0.9695	0.9748	0.9792	0.9970	0.9979
Vision Transformer	0.9815	0.9854	0.9797	0.9834	0.9985	0.9990
ViTCNX	0.9803	0.9907	0.9821	0.9855	0.9985	0.9991

Bold values means the highest score under this metric.