Table 2.

Details of model classification performance evaluation parameters.

Architecture	Accuracy	AUC	95 % CI	Sensitivity	Specificity	F1	Task
Resnet-50	0.853	0.931	0.885–0.977	0.830	0.878	0.854	label-train
	0.864	0.859	0.713–1.000	0.846	0.889	0.880	label-test
Convnext-Base	0.892	0.939	0.891–0.986	0.906	0.878	0.897	label-train
	0.864	0.925	0.844–1.000	0.846	0.889	0.880	label-test
Resnet-101	0.912	0.976	0.952–0.998	0.868	0.959	0.911	label-train
	0.864	0.940	0.876–1.000	0.769	1.000	0.870	label-test
Convnext-V2-Base	0.951	0.991	0.979–1.000	0.925	0.980	0.951	label-train
	0.886	0.955	0.901–1.000	0.808	1.000	0.894	label-test
YOLO-V8	0.853	0.929	0.881–0.976	0.736	0.980	0.839	label-train
	0.841	0.850	0.706–0.994	0.885	0.778	0.868	label-test