Skip to main content

Table 3 Comparison of diagnostic performance of the radiomic model, clinical model, and combined model in the training and internal and external validation cohorts

From: CT whole lung radiomic nomogram: a potential biomarker for lung function evaluation and identification of COPD

Model

AUC (95%CI)

Accuracy (95%CI) (%)

Sensitivity (%)

Specificity (%)

PPV (%)

NPV (%)

Radiomic model

      

 Training cohort

0.888 (0.869 – 0.906)

81.3 (78.9 – 83.4)

85.5

75.2

83.2

78.2

 Internal validation cohort

0.874 (0.844 – 0.904)

78.8 (75.0 – 82.3)

81.1

75.7

81.9

74.7

 External validation cohort

0.846 (0.822 – 0.870)

76.9 (74.3 – 79.4)

54.3

90.6

77.7

76.7

Clinical model

      

 Training cohort

0.732 (0.703 – 0.761)

67.6 (64.9 – 70.2)

65.8

70.1

76.1

58.7

 Internal validation cohort

0.714 (0.669 – 0.759)

65.4 (61.1 – 69.5)

62.5

69.3

73.4

57.6

 External validation cohort

0.777 (0.746 – 0.806)

72.6 (69.9 – 75.3)

62.7

78.5

63.8

77.7

Combined model

      

 Training cohort

0.893 (0.875 – 0.911)

81.7 (79.4 – 83.8)

84.7

77.2

84.3

77.8

 Internal validation cohort

0.873 (0.843 – 0.903)

78.0 (74.2 – 81.5)

81.8

73.1

79.4

76.2

 External validation cohort

0.853 (0.830 – 0.877)

78.3 (75.7 – 80.7)

88.5

61.5

79.2

76.3

  1. AUC area under the curve, CI confidence interval, PPV positive predictive value, NPV negative predictive value