Intratumoral and peritumoral radiomics analysis for preoperative Lauren classification in gastric cancer

Cancer Imaging

Table 2 Performance of different predictive models

Models	Training cohort				Validation cohort
Models	AUC (95% CI)	ACC	SPE	SEN	AUC (95% CI)	ACC	SPE	SEN
Radiomic nomogram	0.745 (0.696–0.795)	0.716	0.659	0.765	0.758 (0.685–0.831)	0.673	0.514	0.793
Combined radiomic signature	0.715 (0.663–0.767)	0.671	0.757	0.598	0.714 (0.636–0.792)	0.642	0.743	0.565
Tumor-based model	0.714 (0.662–0.766)	0.663	0.827	0.525	0.715 (0.637–0.792)	0.630	0.800	0.500
Peripheral ring-based model	0.660 (0.605–0.714)	0.629	0.740	0.534	0.659 (0.576–0.741)	0.617	0.729	0.533
Clinical model 1 (CT_T + CT_N)	0.622 (0.566–0.678)	0.589	0.694	0.500	0.586 (0.498–0.674)	0.574	0.586	0.565
Clinical model 2 (age + CT_T + CT_N)	0.663 (0.608–0.718)	0.623	0.711	0.549	0.605 (0.518–0.692)	0.574	0.586	0.565

NOTE. Abbreviations: AUC Area under the curve, CI Confidence interval, ACC Accuracy, SPE Specificity, SEN Sensitivity, CT_T CT T stage, CT_N CT N stage

ISSN: 1470-7330