Skip to main content

Table 2 Performance of the clinical, radiomics, and combined models

From: Combined model of radiomics, clinical, and imaging features for differentiating focal pneumonia-like lung cancer from pulmonary inflammatory lesions: an exploratory study

Group

AUC (95% CI)

Cut-off

Accuracy

Sensitivity

Specificity

Training cohort (n = 242)

 Clinical model

0.838 (0.788–0.889)

0.626

0.785

0.801

0.760

 Radiomics model

0.804 (0.75–0.858)

0.526

0.740

0.733

0.750

 Combined model

0.915 (0.88–0.95)

0.639

0.847

0.815

0.896

Internal validation cohort (n = 104)

 Clinical model

0.819 (0.738–0.901)

0.740

0.731

0.603

0.902

 Radiomics model

0.877 (0.812–0.942)

0.525

0.833

0.790

0.900

 Combined model

0.899 (0.839–0.96)

0.653

0.833

0.758

0.950

External validation cohort (n = 50)

 Clinical model

0.717 (0.562–0.871)

0.439

0.720

0.852

0.609

 Radiomics model

0.734 (0.59–0.879)

0.364

0.720

0.704

0.739

 Combined model

0.805 (0.681–0.929)

0.283

0.760

0.889

0.609

  1. AUC area under the ROC curve, CI confidence interval