To develop a machine learning model to predict the stone-free rate (SFR) after retrograde intrarenal surgery (RIRS) for lower pole renal stones (LPS), addressing the challenges associated with their anatomical characteristics.
Key Findings:
The SFR for LPS remains lower compared to stones in other locations, indicating a need for improved predictive models.
Acute infundibulopelvic angle and larger preoperative stone size negatively affect SFR, suggesting areas for surgical focus.
Machine learning model can effectively predict SFR based on various patient and stone characteristics, potentially guiding preoperative strategies.
Interpretation:
The machine learning model provides a more accurate prediction of SFR for LPS, aiding in better preoperative decision-making and potentially improving surgical outcomes.
Limitations:
Retrospective design may introduce bias, particularly in patient selection and data collection.
Exclusion of patients with rare conditions limits generalizability, suggesting a need for broader studies.
Conclusion:
Machine learning can enhance the prediction of stone-free rates in patients undergoing RIRS for lower pole renal stones, potentially improving surgical outcomes and informing future research directions.
