Development and Validation of an Imaging-Based Prediction Model for Stone-Free Outcomes after Flexible Ureteroscopy

Abstract

Objective: To devise and externally validate a complete, imaging-based predictive nomogram for estimating the probability of stone-free status (SFS) after flexible ureteroscopy (fURS) for upper urinary tract calculi using advanced parameters from non-contrast computed tomography (NCCT).

Material And Methods: Retrospective study, which comprised 850 consecutive patients with renal and proximal ureteral stones from January 2023 to December 2025. The cohort was randomly divided into a training set (n=600) and an external validation set (n=250). Advanced imaging parameters such as stone volume, maximum Hounsfield units (HU), skin to stone distance (SSD), infundibulopelvic angle (IPA), and renal pelvic area were extracted from pre-operative NCCT. A multivariate logistic regression model was created and presented as a nomogram.

Results: The overall SFS rate was 78.8% (670/850). LASSO and multivariate analyses showed that stone volume, maximal HU, SSD, IPA, and severe hydronephrosis were independent predictors of SFS (all p < 0.05). The imaging-based nomogram demonstrated excellent discrimination in both the training cohort (AUC = 0.865, 95% CI: 0.832-0.898) and the validation cohort (AUC = 0.841, 95% CI: 0.795-0.887).

Conclusion: The imaging-based prediction model newly developed and validated, is an accurate tool for predicting stone-free results following fURS. This will enable the use of advanced NCCT morphological and densitometric parameters for preoperative counseling and for optimizing surgical decision-making in endourology.