Table 5.

Test set

pathological features	model	AUC	95% CI	sensitivity	specificity	Youden index
degree of differentiation	MLP	0.825	0.659–0.967	0.833	0.750	0.583
	LR	0.808	0.649–0.946	0.833	0.725	0.558
	SVM	0.862	0.750–0.967	0.833	0.850	0.683
	DT	0.854	0.700–0.963	0.833	0.875	0.708
	RF	0.858	0.735–0.964	0.833	0.750	0.583
	KNN	0.692	0.519–0.844	0.833	0.450	0.283
T stage	MLP	0.809	0.690–0.905	0.762	0.741	0.503
	LR	0.762	0.633–0.873	0.714	0.630	0.344
	SVM	0.753	0.623–0.857	0.667	0.630	0.297
	DT	0.667	0.543–0.783	0.667	0.667	0.334
	RF	0.727	0.591–0.843	0.714	0.704	0.418
	KNN	0.720	0.586–0.830	0.809	0.407	0.216
N stage	MLP	0.667	0.531–0.799	0.690	0.722	0.412
	LR	0.437	0.294–0.575	0.448	0.444	−0.108
	SVM	0.592	0.435–0.736	0.552	0.500	0.052
	DT	0.723	0.599–0.832	0.724	0.722	0.446
	RF	0.746	0.622–0.872	0.793	0.722	0.515
	KNN	0.560	0.428–0.69	0.621	0.500	0.121