Table 2.

Performance of radiomics signature and the combined model.

		ACC (95% CI)	AUC (95% CI)	IDI (95% CI)	p-value
Primary cohort	Radiomics signature	78.28% (73.30–83.26%)	0.773 (0.701–0.845)	0.102 (0.055–0.150)	2.000e−5
	Combined model	80.54% (76.02–85.07%)	0.803 (0.734–0.872)
Validation cohort	Radiomics signature	80.36% (69.64–89.29%)	0.816 (0.690–0.942)	0.168 (0.032–0.304)	1.516e−2
	Combined model	82.14% (73.21–89.29%)	0.821 (0.697–0.946)

Radiomics signature composed of 6 selected features and constructed by Random Forest method. The combined model incorporated radiomics signature with clinical factors (FIGO stage and age). ACC represents the accuracy of prediction, where ACC describes the percentage of the number of accurately predictive patients and total patients. AUC is the acronym of the area under the curve. 95%CI means a 95% confidence interval. IDI is acronym of Integrated Discrimination Improvement, which is for depicting the difference of two models.