Skip to main content

Table 5 Various model indicators in preoperative diagnosis and XGBoost model

From: An explainable predictive machine learning model of gangrenous cholecystitis based on clinical data: a retrospective single center study

 

Training Set

Test Set

Preoperative diagnosis

XGBoost

Preoperative diagnosis

XGBoost

Accuracy

65.55% (62.54–69.01%)

83.61% (81.38–85.84%)

82.57% (76.15–88.99%)

88.07% (81.65–93.58%)

Sensitivity

51.44% (44.83–58.53%)

87.02% (82.52–91.88%)

92.86% (75.00-100.00%)

71.43% (45.45–93.34%)

Specificity

69.81% (66.37–73.15%)

82.58% (79.82–85.48%)

81.05% (72.91–88.42%)

90.53% (84.54–95.92%)

AUROC

69.81% (56.85–64.58%)

93.38% (91.65–94.91%)

86.95% (78.20–93.30%)

94.40% (88.80-98.39%)

  1. The values were shown as mean (95% confidence interval)
  2. AUROC = area under receiver operating characteristic curve
Back to article page

World Journal of Emergency Surgery

ISSN: 1749-7922