A novel predictive method for URS and laser lithotripsy using machine learning and explainable AI: results from the FLEXOR international database

• DOI: 10.1007/s00345-025-05551-2
• Publication Date: 2025-05-12T08:39:49Z
• Authors: Carlotta Nedbal; Vineet Gauhar; Sairam Adithya; Pietro Tramanzoli; Nithesh Naik; Shilpa Gite; Het Sevalia; Daniele Castellani; Frédéric Panthier; Jeremy Y.C. Teoh; Ben H. Chew; Khi Yung Fong; Mohammed Boulmani; Nariman Gadzhiev; Abhishek Gajendra...; Thomas R. W. Herr...; Olivier Traxer; Bhaskar K. Somani
• Keywords: Tamsulosin; Laser lithotripsy; Ureteroscopy

Abstract Purpose We developed Machine learning (ML) algorithms to predict ureteroscopy (URS) outcomes, offering insights into diagnosis and treatment planning, personalised care improved clinical decision-making. Methods FLEXOR is a large international multicentric database including 6669 patients treated with URS for urolithiasis from 2015 2023. Preoperative postoperative(PO) correlations were investigated through 15 ML-trained algorithms. Outcomes included stone free status (SFS, at 3-month imaging follow up), intraoperative (PCS bleeding, ureteric/PCS injury, need postoperative drainage) PO complications (fever, sepsis, reintervention). ML was applied the prediction, correlation logistic regression analysis. Explainable AI emphasizes key features their contributions output. Results Extra Tree Classifier achieved best accuracy (81%) in predicting SFS. PCS bleed negatively linked ‘positive urine culture‘(-0.08), ‘tamsulosin‘(-0.08), ‘stone location‘(-0.10), ‘fibre optic scope‘(-0.19), ‘Moses Fibre‘(-0.09), ‘TFL‘(-0.09), positively ‘elevated creatine‘(0.25), ‘fever‘(0.11), diameter‘(0.21). ‘PCS injury’ ‘ureteric both showed moderate creatinine‘(0.11), ‘fever‘(0.10), ‘lower pole stone‘(0.09). ‘Tamsulosin‘(0.23) use, presence of ‘multiple‘(0.25) or pole‘(0.25) stones, ‘reusable scope‘(0.17) Fibre’(0.2546) increased risk stent, while ‘digital scope’(-0.13) ‘TFL‘(-0.29) reduced it. ‘Preoperative fever‘(0.10), culture‘(0.16), diameter‘(0.10) may play role ‘PO fever’ ‘sepsis’. SFS mainly influenced by ‘age‘(0.12), ‘preoperative fever‘(0.09), ‘multiple stones‘(0.15), diameter‘(0.17), Fibre“(0.15) ‘TFL‘(-0.28). Conclusion valuable tool accurately outcomes analysing pre-existing datasets. Our model demonstrated strong performance risks laying groundwork development accessible predictive models.