Skip to main content

Table 3 Patient-wise leave-one-out performance of each RFC detection model and the overall performance

From: Detecting localised prostate cancer using radiomic features in PSMA PET and multiparametric MRI for biologically targeted radiation therapy

Patient

PET

mpMRI

PET + mpMRI

Sensitivity

Specificity

ROC-AUC

Sensitivity

Specificity

ROC-AUC

Sensitivity

Specificity

ROC-AUC

11

0.937

0.531

0.681

0.937

0.531

0.885

0.937

0.533

0.878

12

0.662

0.968

0.897

0.877

0.828

0.878

0.663

0.944

0.888

20

0.899

0.555

0.882

0.316

0.823

0.719

0.846

0.745

0.861

26

0.839

0.921

0.938

0.883

0.848

0.935

0.881

0.912

0.946

34

0.865

0.629

0.809

0.150

0.873

0.755

0.798

0.690

0.801

51

0.000

1.000

0.781

0.898

0.843

0.933

0.770

0.898

0.922

53

0.975

0.614

0.892

0.970

0.732

0.931

0.829

0.795

0.899

54

0.936

0.468

0.803

0.827

0.527

0.685

0.888

0.550

0.732

55

0.735

0.666

0.808

0.784

0.818

0.876

0.880

0.726

0.869

56

0.000

1.000

0.674

0.391

0.930

0.771

0.330

0.967

0.780

58

0.849

0.836

0.902

0.730

0.930

0.908

0.833

0.876

0.928

59

0.843

0.750

0.853

0.941

0.703

0.876

0.908

0.709

0.863

60

0.032

0.922

0.636

0.000

0.968

0.806

0.006

0.933

0.607

62

0.915

0.863

0.944

0.754

0.951

0.935

0.891

0.910

0.942

64

0.638

0.817

0.832

0.890

0.739

0.878

0.863

0.768

0.872

65

0.991

0.543

0.864

0.909

0.731

0.919

0.948

0.633

0.911

66

0.623

0.957

0.949

0.515

0.899

0.869

0.614

0.940

0.919

68

0.415

0.880

0.789

0.613

0.830

0.833

0.688

0.838

0.851

Overall

0.781

0.799

0.865

0.802

0.801

0.882

0.842

0.804

0.890