Skip to main content

Table 2 Diagnostic performance of the clinical factors model, the radiomics signature, and the radiomics nomogram

From: MRI-based radiomics nomogram for differentiation of solitary metastasis and solitary primary tumor in the spine

Set

Model

AUC (95% CI)

Sensitivity %

Specificity %

Accuracy %

Training set (n = 98)

Clinical model

0.807 (0.733–0.868)

67 (33/49)

71 (35/49)

69 (68/98)

Radiomics signature

0.975 (0.950–0.993)

96 (47/49)

90 (44/49)

93 (91/98)

Radiomics nomogram

0.980 (0.959–0.995)

96 (47/49)

90 (44/49)

93 (91/98)

Validation set (n = 37)

Clinical model

0.679 (0.533–0.791)

71 (12/17)

60 (12/20)

65 (24/37)

Radiomics signature

0.900 (0.702–0.914)

88 (15/17)

75 (15/20)

81 (30/37)

Radiomics nomogram

0.924 (0.693–0.916)

82 (14/17)

80 (16/20)

81 (30/37)

  1. CI confidence interval, AUC area under the curve