Skip to main content
. 2021 Dec 7;11:737302. doi: 10.3389/fonc.2021.737302

Table 3.

Diagnostic efficacy of four models in the discrimination between low-risk and high-risk GISTs in both training and validation sets.

Model 4 Model 1 Model 2 Model 3
Training set Validation set Training set Validation set Training set Validation set Training set Validation set
Low N = 85 High N = 77 Low N = 34 High N = 35 Low N = 83 High N = 79 Low N = 39 High N = 30 Low N = 96 High N = 66 Low N = 40 High N = 29 Low N = 91 High N = 71 Low N = 32 High N = 37
Accuracy 88.3 88.4 85.8 84.1 86.4 82.6 84.6 82.6
Sensitivity 85.4 88.6 84.2 77.1 76.8 74.3 78.0 85.7
Specificity 91.2 88.2 87.5 91.2 96.2 91.2 91.2 79.4
AUC 0.925 0.897 0.894 0.866 0.909 0.892 0.914 0.892

AUC, area under the receiver operating characteristic curve; Model 4, Clinic + CTU + CTA + CTV; Model 1, Clinic + CTU; Model 2, Clinic + CTA; Model 3, Clinic + CTV; CTU, CT value of unenhanced; CTA, CT value of arterial phase; CTV, CT value of venous phase.