Talyshinskii A. et al., 2025: CT-based radiomics for predicting stone-free rate after ESWL: a systematic review and meta-analysis.

Ali Talyshinskii, Stanislav Ali, Vineet Gauhar, Selcuk Güven, Bhaskar Kumar Somani, Kemal Sarica, Nariman Gadzhiev 
World J Urol. 2025 Sep 20;43(1):566. doi: 10.1007/s00345-025-05933-6. 

Abstract

Purpose: To evaluate the prognostic significance of computed tomography (CT)-based radiomic features of urinary stones in predicting stone-free rate (SFR) after extracorporeal shockwave lithotripsy (ESWL) and laser lithotripsy.

Methods: A systematic literature search of several databases using Boolean operators was performed according to PRISMA guideline and registered in PROSPERO (CRD420250650566). Studies using CT-based radiomics to predict SFR after ESWL or laser lithotripsy and reporting area under the receiver operating characteristic curve (AUC) with 95% confidence intervals (CI) were included. CLAIM and PROBAST checklists were used to assess transparency and and and risk of bias, respectively. A random-effects meta-analysis was performed to pool AUC estimates and assess heterogeneity.

Results: Seven studies were selected and analyzed, of which five described radiomic prediction for ESWL and were analyzed in the meta-analysis. The pooled AUC for radiomics-based models predicting SFR after ESWL was 0.883 (95% CI: 0.840-0.927), with low heterogeneity (I2 = 24.3%) and a prediction interval of 0.77-0.96. All studies were rated at high risk of bias with moderate compliance.

Conclusion: Adding CT-based radiomic features of stones to other clinical and radiologic parameters can predict SFR after ESWL with an AUC of > 0.85. However, the results are based on a small number of studies mostly with a high risk of bias, so follow-up studies performed according to the tenets of using

Comment Peter Alken

All what has to be known is that “the results are based on a small number of studies mostly with a high risk of bias”
See also: Yang R, Zhao D, Ye C, Hu M, Qi X, Li Z. Predicting ESWL success for ureteral stones: a radiomics-based machine learning approach. BMC Med Imaging. 2025 Jul 4;25(1):268. doi: 10.1186/s12880-025-01817-8. PMID: 40615969; PMCID: PMC12228301.
Open access

Peter Alken