Table 2 The diagnostic performance of different models on the internal testing dataset

From: Multimodal model for the diagnosis of biliary atresia based on sonographic images and clinical parameters

Model

AUC

AUPR

Accuracy (%)

Sensitivity (%)

Specificity (%)

P value

Gallbladder model

0.856 (0.765, 0.922)

0.906 (0.884, 0.976)

84.1 (74.8, 91.0)

81.3 (67.4, 91.1)

87.5 (73.2, 95.8)

0.006

Triangular cord model

0.885 (0.800, 0.943)

0.910 (0.873, 0.986)

85.2 (76.1, 91.9)

83.3 (69.8, 92.5)

87.5 (73.2, 95.8)

0.044

Conventional US model

0.948 (0.879, 0.984)

0.970 (0.963, 0.996)

89.8 (81.5, 95.2)

89.6 (77.3, 96.5)

90.0 (76.3, 97.2)

0.539

Gallbladder-clinical model

0.936 (0.864, 0.977)

0.949 (0.899, 0.990)

89.8 (81.5, 95.2)

83.3 (69.8, 92.5)

97.5 (86.8, 99.9)

0.090

Triangular cord-clinical model

0.954 (0.886, 0.987)

0.955 (0.935, 0.995)

88.6 (80.1, 94.4)

85.4 (72.2, 93.9)

92.5 (79.6, 98.4)

0.562

Conventional US-clinical model

0.968 (0.906, 0.994)

0.979 (0.966, 0.996)

89.8 (81.5, 95.2)

91.7 (80.0, 97.7)

87.5 (73.2, 95.8)

–

  1. 95% confidence intervals are included in parentheses
  2. US ultrasound, AUC area under receiver operating characteristic curve, AUPR area under Precision-Recall curve
  3. The P values were from the comparison between the AUC of the conventional US-clinical model and the AUCs of the other models. Differences between various AUCs were compared using a Delong test.
Back to article page