A radiomics signature derived from CT imaging to predict MSI status and immunotherapy outcomes in gastric cancer: a multi-cohort study

Table 4 The specific performances of models for evaluate MSI status

Model	Cohort	AUC	Accuracy	Sensitivity	Specificity
Clinical	Training	0.681 (0.581, 0.781)	0.690	0.709	0.625
	Validation	0.753 (0.613, 0.893)	0.763	0.804	0.615
Radiomics signature	Training	0.851 (0.782, 0.919)	0.683	0.938	0.609
	Validation	0.816 (0.706, 0.926)	0.695	1	0.608
Combined	Training	0.875 (0.815, 0.935)	0.704	1	0.618
	Validation	0.799 (0.659, 0.940)	0.780	0.804	0.692

Note: Data in parentheses are 95% CI. MSI = microsatellite instability, AUC = area under the receiver operating characteristic curve, CI = confidence intervals

ISSN: 1471-2407