To apply machine learning approaches for preoperative risk stratification of post-PCNL SIRS and to explore the combined effects of sarcopenia and staghorn stones.
Approach:
Study Design: A retrospective study involving 755 patients who underwent PCNL, with data collected from electronic medical records and categorized into clinical, laboratory, and imaging domains.
Data Analysis: Patients were randomly allocated to training and validation sets, and machine learning models were developed to predict postoperative SIRS.
Sarcopenia Assessment: Skeletal muscle index was calculated from preoperative CT images to evaluate sarcopenia.
Key Findings:
Sarcopenia and staghorn calculi are potential predictors of postoperative SIRS following PCNL.
Machine learning models can enhance the accuracy of preoperative risk stratification for SIRS.
Interpretation:
The study assesses the relationship between sarcopenia, staghorn calculi, and postoperative complications after PCNL.
Limitations:
Retrospective design may introduce selection bias.
Findings may not be generalizable to all populations.
Conclusion:
The study aims to improve early identification and intervention strategies for SIRS in patients undergoing PCNL.
