Prediction of lymphovascular space invasion using a combination of tenascin-C, cox-2, and PET/CT radiomics in patients with early-stage cervical squamous cell carcinoma

Table 4 Performance of each model in the training and testing datasets

Model	Training dataset				Testing dataset				External dataset
Model	AUC (95% CI)	p	Sen	Spe	AUC (95% CI)	p	Sen	Spe	AUC (95% CI)	p	Sen	Spe
Radiomics model	0.914 (0.814–0.970)	< 0.001	0.93	0.74	0.756 (0.545–0.904)	0.014	0.58	0.92	0.806 (0.606–0.933)	< 0.001	0.80	0.82
Protein model	0.756 (0.630–0.857)	< 0.001	0.90	0.48	0.631 (0.417–0.814)	0.238	0.50	0.77	0.503 (0.302–0.703)	0.979	0.87	0.27
Combined model	0.905 (0.803–0.965)	< 0.001	1.00	0.64	0.801 (0.594–0.932)	0.003	0.67	1.00	0.782 (0.577–0.918)	0.002	0.53	1.00

AUC area under the curve, CI confidence interval, Sen sensitivity, Spe specificity

ISSN: 1471-2407