Skip to main content

Table 2 Predictive performance of radiomic signature, CPR model and radiomic nomogram

From: DWI-based radiomic signature: potential role for individualized adjuvant chemotherapy in intrahepatic cholangiocarcinoma after partial hepatectomy

Models

AUC (95%CI)

ACC

SEN

SPE

P

Training set

     

CPR model

0.697 (0.592–0.802)

0.644

0.655

0.625

–

Radiomic signature

0.823 (0.729–0.917)

0.747

0.691

0.844

0.06a

Radiomic nomogram

0.876 (0.796–0.955)

0.851

0.855

0.844

0.001b

Validation set

     

CPR model

0.621 (0.434–0.808)

0.649

0.773

0.467

–

Radiomic signature

0.753 (0.597–0.909)

0.676

0.591

0.800

0.274a

Radiomic nomogram

0.821 (0.684–0.959)

0.757

0.773

0.733

0.01b

  1. p values < 0.05 were considered statistically significant and are shown in bolded font
  2. AUC the area under the receiver operating characteristic curve, 95%CI 95% confidence intervals, ACC accuracy, SEN sensitivity, SPE specificity, CPR model clinicopathological and MR radiographic model
  3. aDeLong test was used to compare the difference of AUC between radiomic signature and CPR model
  4. bDeLong test was used to compare the difference of AUC between radiomic nomogram and CPR model