Table 3 AUC values with 95%CIs for the prediction of metastasis based on 3 models in stratified training and test set

From: Predictive models of melanoma metastasis based on dermatoscopy in an international retrospective human reader study

Training set

Model 1

Model 2

Model 3

Fold 1

0.784 (0.694–0.874)

0.699 (0.587–0.811)

0.803 (0.720–0.888)

Fold 2

0.765 (0.646–0.884)

0.766 (0.656–0.876)

0.822 (0.711–0.932)

Fold 3

0.906 (0.830–0.981)

0.821 (0.719–0.922)

0.908 (0.843–0.972)

Fold 4

0.795 (0.700–0.890)

0.804 (0.703–0.904)

0.857 (0.774–0.941)

Fold 5

0.760 (0.650–0.870)

0.696 (0.575–0.817)

0.751 (0.638–0.864)

Average

0.802 (0.752–0.854)

0.758 (0.707–0.801)

0.824 (0.784–0.877)

DeLong’s test

p > 0.05 (vs Model 2)

p > 0.05 (vs Model 3)

p > 0.05 (vs Model 1)

Test set

Model 1

Model 2

Model 3

 

0.814 (0.732–0.896)

0.772 (0.682–0.870)

0.834 (0.757–0.912)

DeLong’s test

p > 0.05 (vs Model 2)

p > 0.05 (vs Model 3)

p > 0.05 (vs Model 1)

  1. AUC area under the curve, CI confidence interval. Two-sided p-values derived from DeLong’s test and pairwise comparisons of AUC values between 3 models were conducted. Exact p-values are provided below (training set: M1/M2 p = 0.20, M1/M3 p = 0.52, M2/M3 p = 0.10, test set: M1/M2 p = 0.51, M1/M3 p = 0.72, M2/M3 p = 0.31). Correction for multiple comparisons was not applied.
Back to article page