Skip to main content

Table 4 Diagnostic performance of the clinical model, RS and RN

From: A CT-based radiomics nomogram for predicting histologic grade and outcome in chondrosarcoma

Model

AUC (95%CI)

Sensitivity %

Specificity %

Accuracy %

Training set (n = 139)

    

Clinical model

0.776 (0.696–0.855)

72.900

76.300

74.800

RS

0.843 (0.778–0.908)

57.600

93.800

78.400

RN

0.890 (0.839–0.941)

64.400

88.800

78.400

Validation set (n = 57)

    

Clinical model

0.705 (0.563–0.847)

58.300

75.800

68.400

RS

0.835 (0.724–0.945)

62.500

90.900

78.900

RN

0.842 (0.727–0.958)

66.700

90.900

80.700

  1. CI: confidence interval; Data in the parentheses are raw data
  2. RS: radiomics signature
  3. RN: radiomics nomogram