Table 3.

Diagnostic performance of individualized prediction models.

		AUC (95% CI)	ACC	SEN	SPE	PPV	NPV	Cutoff
Training cohort (n=234)
Radiomics	Logistic	0.847(0.796-0.898)	0.791	0.809	0.783	0.604	0.909	0.287
	Tree	0.798(0.737-0.858)	0.786	0.765	0.795	0.605	0.892	0.210
	SVM	0.847(0.796-0.898)	0.791	0.809	0.783	0.604	0.909	0.282
Clinical		0.775(0.709-0.841)	0.752	0.691	0.777	0.560	0.860	0.309
Comb		0.876(0.828-0.924)	0.816	0.779	0.831	0.654	0.902	0.275
Testing cohort (n=100)
Radiomics	Logistic	0.826(0.733-0.919)	0.760	0.679	0.792	0.559	0.864	0.284
	Tree	0.696(0.591-0.801)	0.730	0.643	0.764	0.514	0.846	0.200
	SVM	0.826(0.733-0.919)	0.760	0.679	0.792	0.559	0.864	0.281
Clinical		0.798(0.707-0.890)	0.650	0.607	0.667	0.415	0.814	0.300
Comb		0.867(0.792-0.941)	0.810	0.714	0.847	0.645	0.884	0.277

Logistic, logistic regression; Tree, decision tree; SVM, support vector machine; AUC, area under the curve; CI, confidence interval; ACC, Accuracy; SEN, Sensitivity; SPE, specificity; PPV, positive predictive value; NPV, negative predictive value. Radiomics, radiomics model; Clinical, clinical model; Comb, combined model.