Scholarlink

<p>To develop a machine-learning-based predictive model for postoperative urinary sepsis following ureteroscopic lithotripsy, providing a scientific basis for early clinical identification of high-risk patients. A total of 927 patients who underwent ureteroscopic lithotripsy at a Grade III-A hospital in Guizhou Province from September 2024 to September 2025 were enrolled as the study subjects. Clinical data were collected and randomly divided into a training set (70%) and a test set (30%) according to a 7:3 ratio. Four machine-learning algorithms—SVM, LR, RF, and XGBoost—were used to construct predictive models. The performance of each model was comprehensively evaluated using metrics such as the area under the receiver operating characteristic curve (AUC), accuracy, precision, and F1 score. The clinical utility of each model was further assessed using decision curve analysis (DCA) and calibration curves to identify the optimal model. The SHAP method was then employed to analyze the contribution of each variable in the optimal model. The AUC values for the four models—SVM, LR, RF, and XGBoost—were 0.96 (0.93–0.98), 0.95 (0.91–0.98), 0.98 (0.96–1.00), and 0.98 (0.97–1.00), respectively, with the XGBoost model demonstrating the best performance. Variable importance analysis based on the optimal model revealed that the top six key predictors were procalcitonin, albumin, degree of hydronephrosis,5-frailty score, maximum stone diameter, and urinary tract infection.Machine-learning models can effectively predict the risk of postoperative urinary sepsis following ureteroscopic lithotripsy, with the XGBoost model performing the best. The key variables—including patient functional status, stone characteristics, and immune indicators—provide a scientific basis for early identification of uroseptic sepsis and offer a quantifiable decision-making tool for precise prevention and management of postoperative complications.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Development of a machine learning model for predicting urosepsis after ureteroscopic lithotripsy

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年3月31日

查看全文
  • Aobing Mei,
  • Sudan Zeng,
  • Wen Zhou,
  • Miaoqing Cai,
  • Jiamei Wang,
  • Xu Sun,
  • Mei Chen

摘要

To develop a machine-learning-based predictive model for postoperative urinary sepsis following ureteroscopic lithotripsy, providing a scientific basis for early clinical identification of high-risk patients. A total of 927 patients who underwent ureteroscopic lithotripsy at a Grade III-A hospital in Guizhou Province from September 2024 to September 2025 were enrolled as the study subjects. Clinical data were collected and randomly divided into a training set (70%) and a test set (30%) according to a 7:3 ratio. Four machine-learning algorithms—SVM, LR, RF, and XGBoost—were used to construct predictive models. The performance of each model was comprehensively evaluated using metrics such as the area under the receiver operating characteristic curve (AUC), accuracy, precision, and F1 score. The clinical utility of each model was further assessed using decision curve analysis (DCA) and calibration curves to identify the optimal model. The SHAP method was then employed to analyze the contribution of each variable in the optimal model. The AUC values for the four models—SVM, LR, RF, and XGBoost—were 0.96 (0.93–0.98), 0.95 (0.91–0.98), 0.98 (0.96–1.00), and 0.98 (0.97–1.00), respectively, with the XGBoost model demonstrating the best performance. Variable importance analysis based on the optimal model revealed that the top six key predictors were procalcitonin, albumin, degree of hydronephrosis,5-frailty score, maximum stone diameter, and urinary tract infection.Machine-learning models can effectively predict the risk of postoperative urinary sepsis following ureteroscopic lithotripsy, with the XGBoost model performing the best. The key variables—including patient functional status, stone characteristics, and immune indicators—provide a scientific basis for early identification of uroseptic sepsis and offer a quantifiable decision-making tool for precise prevention and management of postoperative complications.