Skip to main content

Table 2 Diagnostic Performance of the clinical factors model, the radiomics signature and the radiomics nomogram for detection of focal nodular hyperplasia

From: A CT-based radiomics nomogram for differentiation of focal nodular hyperplasia from hepatocellular carcinoma in the non-cirrhotic liver

Model

Training set (n = 119)

Validation set (n = 37)

Cutoff

AUC (95% CI)

Sensitivity*

Specificity*

Accuracy*

Cutoff

AUC (95% CI)

Sensitivity*

Specificity*

Accuracy*

Clinical factors model

−1.513

0.799(0.719,0.879)

83.33(35/42)

70.13(54/77)

74.79(89/119)

−0.434

0.769(0.590,0.949)

69.23(9/13)

87.50(21/24)

81.08(30/37)

Radiomics signature

0.543

0.964(0.934,0.995)

88.10(37/42)

94.81(73/77)

92.44(110/119)

−0.306

0.865(0.725,1.000)

76.92(10/13)

91.67(22/24)

86.49(32/37)

Radiomics nomogram

−1.405

0.979(0.959,0.998)

92.86(39/42)

92.21(71/77)

92.44(110/119)

−2.795

0.917(0.800,1.000)

92.31(12/13)

87.50(21/24)

89.19(33/37)

  1. Note: CI (Confidence interval); * Numbers in parentheses were used to calculate percentages