Skip to main content

Table 3 The summaries of performance of different predictive models with radiomics and nomogram in both the training and validation cohort on MR images

From: MR-based radiomics-clinical nomogram in epithelial ovarian tumor prognosis prediction: tumor body texture analysis across various acquisition protocols

Characteristics

Training AUC

(95% CI)

Validation AUC

(95% CI)

TP

TN

FP

FN

ACC

SEN

SPE

PPV

NPV

Clinical

0.704 (0.619ā€“0.787)

0.685 (0.545ā€“0.825)

17

17

20

2

0.607

0.895

0.459

0.459

0.895

T1WI

signature

0.845 (0.771ā€“0.906)

0.553 (0.382ā€“0.736)

10

23

14

9

0.589

0.526

0.622

0.417

0.719

CE T1WI

signature

0.837 (0.755ā€“0.910)

0.593 (0.413ā€“0.770)

7

30

7

12

0.661

0.368

0.811

0.500

0.714

DWI

signature

0.848 (0.783ā€“0.909)

0.603 (0.441ā€“0.765)

7

15

22

12

0.393

0.368

0.405

0.241

0.556

T2WI

signature

0.844 (0.762ā€“0.917)

0.771 (0.629ā€“0.894)

10

31

6

9

0.732

0.526

0.838

0.625

0.775

T1WI

nomogram

0.855 (0.794ā€“0.910)

0.724 (0.587ā€“0.865)

14

24

13

5

0.679

0.737

0.649

0.519

0.828

CE T1WI

nomogram

0.868 (0.813ā€“0.918)

0.702 (0.557ā€“0.849)

12

24

13

7

0.643

0.632

0.649

0.480

0.774

DWI

nomogram

0.767 (0.681ā€“0.850)

0.727 (0.576ā€“0.870)

13

27

10

6

0.714

0.684

0.730

0.565

0.818

T2WI-3D

nomogram

0.866 (0.792ā€“0.931)

0.818 (0.691ā€“0.932)

10

33

4

9

0.768

0.526

0.892

0.714

0.786

T2WI-2D

nomogram

0.830 (0.765ā€“0.890)

0.720 (0.559ā€“0.873)

13

25

12

6

0.679

0.684

0.676

0.520

0.806

  1. TP True positive, TN True negative, FP False positive, FN False negative, ACC Accuracy, SEN Sensitivity, SPE Specitivity, PPV Positive predictive value, NPV Negative predictive value