{"paper_id":"0e39063d-fb72-409b-a5de-c86dcd29d78a","body_text":"Seeing Beyond the Scan: Predicting ESWL Success Through Quantitative CT Parameters | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Seeing Beyond the Scan: Predicting ESWL Success Through Quantitative CT Parameters Hyangmi Kim, Mijeong Kim, Yechan Goo, Jung Ki Jo This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-6979264/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Purpose Extracorporeal shock wave lithotripsy (ESWL) is a non-invasive treatment for urolithiasis; however, predicting its success remains challenging. This study aimed to provide guidance on the key predictors and clinical criteria for anticipating the success of ESWL. Materials and Methods We retrospectively analyzed 488 patients who underwent ESWL between 2011 and 2020. Predictive features included patient demographics, stone characteristics on non-contrast computed tomography (NCCT), composite scores (triple D, quadruple D), and ESWL energy parameters. Logistic regression, random forest, and decision tree models were used to identify key predictors of treatment success based on accuracy, sensitivity, and specificity. Results Logistic regression showed 94.3% accuracy (area under the curve [AUC] 0.883), with 92.6% (AUC 0.936) for random forest and 91.4% (AUC 0.559) for decision tree. Small stones, lower Hounsfield unit, higher stone heterogeneity index, and favorable locations (pelvis/upper) were consistent predictors of success. Although the sample size was robust, the single-center design may have limited the external validity of the findings. Conclusions NCCT-derived stone characteristics are sufficient to predict ESWL outcomes. This study provides a threshold-based guidance for each predictor, enabling clinicians to make timely and informed decisions regarding ESWL in patients with urolithiasis. Clinical decision support Extracorporeal shock wave lithotripsy (ESWL) Machine learning Non-contrast computed tomography (NCCT) Urolithiasis Figures Figure 1 Figure 2 Figure 3 I. Introduction Urolithiasis affects approximately 9% of the U.S. population, with 600,000 new cases reported annually. [ 1 ] In South Korea, the number of treated cases increased from 301,000 in 2017 to 317,472 in 2022. [ 2 ] Clinical management depends on stone size, location, and composition, ranging from observation to pharmacological dissolution, extracorporeal shock wave lithotripsy (ESWL), or surgery. Very small stones (≤ 4 mm) may spontaneously pass through without causing severe symptoms or renal dysfunction. As the incidence of urolithiasis continues to increase, accurate and timely treatment selection is becoming increasingly important to minimize patient burden and optimize clinical outcomes. ESWL is a noninvasive procedure using external shock waves, preferred for treating small stones that are unlikely to pass naturally. It offers minimal pain, outpatient treatment, rapid recovery, and high success rates (80–90%). Treatment success is influenced by factors related to the stone (e.g., size, density), the patient (e.g., sex, age, and renal function), and ESWL treatment itself (e.g., energy escalation). [ 3 ] Although prior studies have explored factors, most have focused on identifying significant variables without providing guidance on prioritization or proposing standardized cut-off values to aid in patient selection. Moreover, traditional statistical methods often fail to fully capture the interactions between these variables, limiting their clinical applicability. To address these gaps, this study aims to systematically rank the key predictors of ESWL success using machine learning models and propose clinically meaningful cut-off values supporting stepwise, evidence-based decisions. By bridging predictive modeling and clinical utility, our approach seeks to provide a practical algorithm for clinicians when determining ESWL suitability. II. Methods This study was approved by the Institutional Review Board of H Hospital, South Korea (IRB No. 2022-08-017). We retrospectively analyzed the data of 488 patients who visited H Hospital for urolithiasis between 2011 and 2020. The dataset included blood and urine test results, computed tomography (CT) imaging findings, and treatment outcomes following ESWL. To predict ESWL success rates, we applied logistic regression, random forest, and decision tree models based on stone-related factors (volume, location, Hounsfield unit (HU), stone heterogeneity index (SHI), skin-to-stone distance (SSD), triple D, and quadruple D), patient characteristics (age and sex), and treatment-related variables (shock intensity and delivery efficiency). Stone Volume. Smaller stones are consistently associated with higher stone-free rates following a single ESWL session and are recognized as independent predictors of treatment success. [ 3 – 5 ] The European Association of Urology (EAU) Guidelines (2023) recommend ESWL as first-line therapy for stones ≤ 20 mm, with optimal outcomes for stones ≤ 10 mm. Prior studies reported success rates of approximately 85% for stone volumes ≤ 150 mm³. [ 3 ] Stone Location. Stones located in the upper and mid-renal calyces tend to respond better to ESWL, whereas lower pole stones are less likely to be successfully cleared. ESWL is generally recommended for proximal ureteral stones < 10 mm, with reported success rates around 70%. [ 4 – 5 ] HU. Stone fragility is correlated with density. Hard stones, such as cystine or calcium oxalate monohydrate, showed higher HU and lower ESWL success rates. [ 6 ] Stones with HU > 1000 are associated with higher failure risks, whereas those with HU ≤ 1000 demonstrate significantly higher success. [ 4 , 7 – 9 ] SHI. The SHI quantifies the internal variability of stone. A higher SHI indicates greater structural heterogeneity, facilitating shockwave fragmentation and improving ESWL success. [ 10 , 11 ] SSD. An SSD of ≤ 10 cm has been associated with better ESWL outcomes. [ 4 , 12 – 14 ] However, recent studies suggest that stone-related characteristics may be more predictive of treatment success than patient body habitus alone. [ 15 ] Triple D Score. The triple D Score integrates stone volume, HU, and SSD to predict the likelihood of ESWL success. [ 3 , 16 ] Higher scores have been correlated with improved outcomes. In this study, we adopted cut-off values based on previous research: stone volume ≤ 150 mm³, SSD ≤ 9 cm, and HU ≤ 1000. Quadruple D Score. The quadruple D score extends the triple D score by incorporating lower pole stone (LPS) status. [ 17 ] It uses the criteria of stone volume ≤ 150 mm³, SSD ≤ 9 cm, HU ≤ 1000, and absence of LPS, with higher scores being associated with improved ESWL success rates. [ 11 ] In the present study, the cut-off point for the quadruple D score was based on that of the triple D score, with the addition of stone location (lower). Sex. Females may have slightly higher ESWL success rates than males [ 7 ]; however, sex is not considered an independent predictor when compared to factors such as stone size, location, or density. [ 9 ] Age. Younger patients tend to achieve higher ESWL success rates because of their more efficient renal and ureteral function. [ 18 ] In contrast, older patients may show worse outcomes owing to anatomical changes and harder stone compositions. Shock Wave (Energy) Intensity and Efficiency. Gradual escalation of the shock wave intensity has been shown to enhance stone fragmentation while minimizing renal injury. [ 5 ] Recent studies have also emphasized the importance of efficiency metrics, such as the ratio of delivered energy to stone volume (SMLI/SV). [ 3 ] We used Python (version 3.10) and the scikit-learn library (version 1.2.2) for data preprocessing and machine learning-based predictive modeling (logistic regression, random forest, and decision tree), based on the aforementioned variables. III. Results A total of 488 patients (345 males and 143 females) were included in the study, with an average age of 50 years; 73.16% were aged > 40 years. In addition, there were 383 successful cases of stone removal following extracorporeal shock wave lithotripsy (ESWL) and 105 cases in which the treatment failed. Data pre-processing involved random forest imputation for categorical variables and median replacement for continuous variables. Feature engineering created three derived variables: the total ESWL energy per stone volume, HU, and SSD. Three predictive models—logistic regression, random forest, and decision tree—were developed to compare the performance and identify key predictors. Prior to analysis, the distribution of eight continuous variables, excluding the derived variables, was examined in relation to ESWL outcomes (success vs. failure), as illustrated, in Figure 1. The distributions of age, total ESWL energy, stone volume, HU, SHI, and SSD appeared similar between the success and failure groups, which contrasts with findings from previous studies. However, because ESWL success may depend on subtle differences in variable values, we proceeded to identify significant predictors using logistic regression, random forest, and decision tree analyses. 3.1. Logistic Regression Logistic regression was initially employed as a baseline model because of its ability to statistically evaluate the effect of each variable on ESWL success and maintain consistency with prior studies. The model demonstrated a strong performance, achieving an accuracy of 94.3% and an area under the curve (AUC) of 0.883. Sensitivity was notably high at 98.9% (365/369 successful cases correctly classified), indicating excellent identification of likely successful candidates. However, specificity was relatively low at 44.1% (15/34 failures correctly classified), reflecting limited ability to detect treatment failures. These results highlight the model’s effectiveness in reducing missed treatment opportunities, which is a key priority in clinical decision-making. The significant predictors of ESWL success included stone location in the upper ureter (odds ratio [OR] = 9.81, p < 0.001), renal pelvis (OR = 6.47, p = 0.01), higher SHI (OR = 1.02, p < 0.001), and lower HU (OR = 0.99, p < 0.001). Notably, in contrast to previous studies, a greater SSD was positively associated with ESWL success (OR = 1.03, p = 0.01). A higher total ESWL energy per unit stone volume was associated with lower success rates, suggesting that the efficiency of energy delivery may be more important than the absolute energy applied. No significant associations were found for age, sex, stone volume, total ESWL energy, or other derived variables. Logistic regression analysis is valued for its interpretability and ability to support statistical inference. However, it has limitations in capturing complex nonlinear relationships between variables and is sensitive to multicollinearity, which can reduce its predictive performance compared to tree-based models. To address these limitations and enhance predictive accuracy, we utilized the random forest and decision tree methods. 3.2. Random Forest Model As mentioned earlier, the random forest model was implemented to capture complex nonlinear relationships, which is a limitation of logistic regression, and to improve model stability. The data were split into an 80% training dataset and a 20% test dataset, using 100 decision trees (n_estimators = 100). The random forest analysis showed that the model had an accuracy of 92.6%, a sensitivity of 98.6%, a specificity of 28.6%, and an AUC of 0.93, which was better than that of the logistic regression model (AUC = 0.883). This high sensitivity was clinically meaningful, as it minimized the risk of missing patients who could be successfully treated with ESWL. The results of the variable importance analysis identified stone volume, HU, and SHI as the most influential predictors of ESWL success. Figure 2 presents the relative importance of these variables as determined by the random forest model. Although these results were not completely consistent with the results of the logistic regression analysis, some of the significant variables were consistent with the results of previous studies and the logistic regression analysis. Stone volume, HU, and SHI were reliable predictors of ESWL success. 3.3. Decision Tree Model Although logistic regression and random forest analyses demonstrated strong predictive performance, clearer criteria are needed for clinical decision-making by healthcare professionals. Therefore, using the same dataset, we implemented a decision tree model with a maximum depth of four to prevent overfitting and enhance visualization. The model was built based on the key variables and criteria that affect ESWL success for urolithiasis. It presented clear decision rules and classification paths, providing an intuitive understanding for clinicians. The decision tree achieved an accuracy of 91.4%, a sensitivity of 95.9%, a specificity of 42.9%, and an AUC of 0.559. Although its overall performance was lower than that of logistic regression and random forest, its high sensitivity highlights its potential value in supporting clinical decisions. The decision tree model offers strong explanatory power by capturing nonlinear relationships and visually representing the decision rules (Figure 3). The primary classification factor was stone volume (≤ 27.99mm³), which was significantly associated with a higher ESWL success rate, identifying smaller stones as a major predictor. Further stratification based on stone size, location, and SHI revealed that among patients with clearly defined stone locations, those with SHI ≥ 50 had a higher success rate, suggesting that the greater internal heterogeneity leads to better fragmentation. For stones with volumes ≥ 27.99 mm³, HU became the next classification criterion. Stones with HU ≥ 393.61 (mean HU 486.66 in the success group) and SHI ≤ 256.18 had a higher failure rate, indicating that more homogeneous stone material, is harder to fragment with ESWL. These findings highlight the utility of decision trees as interpretable tools that provide intuitive predictive paths to guide clinical decision-making for ESWL. In summary, we employed three machine learning approaches (logistic regression, random forest, and decision trees) to predict ESWL success based on features obtained from NCCT, patient demographic characteristics, and treatment variables. Model performance was evaluated using standard classification metrics: accuracy, sensitivity, specificity, and AUC. The use of three approaches was intentional to leverage the strengths of each method. Logistic regression was chosen to allow comparison with previous studies through statistical inference, and is served as a baseline model for the random forest and decision tree models. The random forest model facilitated robust modeling of nonlinear relationships, while the decision tree provided an interpretable, rule-based decision path that enhances clinical applicability. This comparative approach, using multiple models, was designed to strengthen predictive validity and ensure practical relevance for real-world clinical decision-making. IV. Discussion ESWL is a noninvasive procedure that uses externally generated shock waves to fragment urinary stones. It is widely used for treating renal and ureteral stones, particularly small ones, without requiring hospitalization, anesthesia, or surgery. ESWL remains an effective first-line treatment for appropriately selected patients. [ 3 , 9 , 19 – 20 ] For small stones, repeated ESWL can achieve stone-free rates of up to 90%. [ 21 ] Identifying patients most likely to benefit is clinically important. Various predictive models have evaluated factors associated with success. Prior studies consistently show that smaller stone size, [ 3 , 4 , 8 , 13 , 22 – 24 ] lower HU values, [ 6 ] greater heterogeneity, [ 25 ] and favorable locations (renal pelvis or upper ureter) [ 7 , 22 – 24 , 26 – 27 ] predict higher success. However traditional statistical methods may not fully capture complex interactions between variables. [ 3 , 13 , 22 , 26 ] To address this, we applied three machine learning models—logistic regression, random forest, and decision tree—to predict ESWL outcomes. Consistent with prior studies, this study confirmed that a small stone volume (≤ 27.99 mm³), HU (< 393.61), SHI (≥ 50), and stone location (renal pelvis or upper ureter) were key predictors of ESWL success in order of importance. Notably, SSD showed a positive correlation with ESWL success, contrasting with prior studies, but this alone may not reliably predict outcomes due to potential confounding factors in the logistic regression model. Although both the logistic regression and random forest models performed well, they lacked explainability for clinical decision-making. Therefore, we developed a decision tree model to visualize the classification process and threshold for key predictors. Based on thresholds from our models, we propose a stepwise clinical decision-making algorithm for ESWL candidacy. There thresholds can be assessed on routine NCCT scans, enabling clinician to stratify patients and select appropriate treatment. Integrating these quantitative criteria into the pre-treatment evaluation can enhance ESWL decision-making and outcomes. This study has several limitations. First, the single-center retrospective design may limit generalizability, underscoring the need for validation in external cohorts. Second, the small number of failure cases may have reduced the model’s specificity; therefore, future studies should consider using oversampling or data balancing techniques to address this issue. Third, although AUC comparisons were conducted, testing (e.g., the DeLong test) should be applied in future studies to confirm significant differences in model performance and strengthen the robustness of validation. Further research should validate the findings of this study in a larger patient population to enhance the generalizability of the results. Additionally, studies should investigate how to integrate this system into real-world clinical workflows to support real-time decision-making. V. Conclusion In this study, we evaluated three machine learning models to predict the success rate of ESWL and found that a small stone volume, small HU, large SHI, and stone location (renal pelvis or upper ureter) were the key predictors. As these features can be easily obtained from NCCT, they serve as practical indicators to help clinicians make timely and effective decisions regarding ESWL procedures, thereby supporting more efficient clinical management. Declarations Ethics approval and consent to participate This retrospective study was approved by the Institutional Review Board of Hanyang University Medical Center (IRB No. 2022-08-017). The requirement for informed consent was waived due to the retrospective nature of the study. Consent for publication Not applicable Funding No funding was received for conducting this study. Competing interests The authors declare that they have no competing interests. Author Contribution H.M.K. designed the study, conducted data analysis, and drafted the manuscript. Y.C.G. collected the data. M.J.K. contributed to data collection and manuscript editing. J.K.J. supervised the entire research process as the senior advisor and provided overall guidance throughout the study. All authors reviewed and approved the final version of the manuscript. References Ureteroscope mark size share COVID. Fortune Business Insights . Published May 19, 2025. https://www.fortunebusinessinsights.com/ureteroscope-market-103455. Accessed March 3, 2026. Health Insurance Review & Assessment Service. Report on statistics of diseases and medical procedures in daily life. 2023. Snicorius M, Drevinskaitė M, Miglinas M, et al. A prospective study on the impact of clinical factors and adjusted Triple D system for success rate of ESWL. Medicina (Kaunas) . 2023;59:1827. doi.org/10.3390/medicina59101827 Bajaj M, Smith R, Rice M, Zargar-Shoshtari K. Predictors of success following extracorporeal shock-wave lithotripsy in a contemporary cohort. Urol Ann . 2021;13:282–7. doi.org/10.4103/UA.UA_155_19 Ordon M, Andonian S, Blew B, Schuler T, Chew B, Pace KT. CUA guideline. CUA Guideline: management of ureteral calculi. Can Urol Assoc J . 2015;9:E837–51. doi.org/10.5489/cuaj.3483 Yoon JH, Park S, Kim SC, Park S, Moon KH, Cheon SH, et al. Outcomes of extracorporeal shock wave lithotripsy for ureteral stones according to ESWL intensity. Transl Androl Urol. 2021;10:1588–95. doi.org/10.21037/tau-20-1397 El-Nahas AR, El-Assmy AM, Mansour O, Sheir KZ. A prospective multivariate analysis of factors predicting stone disintegration by extracorporeal shock wave lithotripsy: the value of high-resolution noncontrast computed tomography. Eur Urol . 2007;51:1688–93; discussion 1693. doi.org/10.1016/j.eururo.2006.11.048 Ouzaid I, Al-Qahtani S, Dominique S, Hupertan V, Fernandez P, Hermieu JF, et al. A 970 Hounsfield units (HU) threshold of kidney stone density on non-contrast computed tomography (NCCT) improves patients’ selection for extracorporeal shockwave lithotripsy (ESWL): evidence from a prospective study. BJU Int . 2012;110;110:E438–42. doi.org/10.1111/j.1464-410X.2012.10964.x Türk C, Petřík A, Sarica K, et al. EAU guidelines on interventional treatment for urolithiasis. Eur Urol . 2016;69:475–82. doi.org/10.1016/j.eururo.2015.07.041 Iqbal N, Hasan A, Nazar A, et al. Role of stone heterogeneity index in determining success of shock wave lithotripsy in urinary calculi. J Clin Transl Res . 2021;7:241–7. Şendoğan F, Bulut M, Çanakcı C, et al. Quadruple-D score in the success rate of extracorporeal shock wave lithotripsy of renal stones in pediatric population. Urolithiasis . 2024;52:163. doi.org/10.1007/s00240-024-01657-1 Cho KS, Jung HD, Ham WS, et al. Optimal skin-to-stone distance is a positive predictor for successful outcomes in upper ureter calculi following extracorporeal shock wave lithotripsy: a Bayesian model averaging approach. PLOS One . 2015;10:e0144912. doi.org/10.1371/journal.pone.0144912 Pareek G, Armenakas NA, Panagopoulos G, Bruno JJ, Fracchia JA. Extracorporeal shock wave lithotripsy success based on body mass index and Hounsfield units. Urology. 2005;65:33–6. doi.org/10.1016/j.urology.2004.08.004 Patel T, Kozakowski K, Hruby G, Gupta M. Skin to stone distance is an independent predictor of stone-free status following shock wave lithotripsy. J Endourol . 2009;23:1383–5. doi.org/10.1089/end.2009.0394. Çanakcı C, Dinçer E, Şimşek B, et al. Effect of tissue densities at the skin-to-stone distance on the success of shockwave lithotripsy. J Urol Surg . 2024;11:14–8. doi.org/10.4274/jus.galenos.2023.2023.0010 Tran TY, McGillen K, Cone EB, Pareek G. Triple D score is a reportable predictor of shockwave lithotripsy stone-free rates. J Endourol . 2015;29:226–30. doi.org/10.1089/end.2014.0212 Ichiyanagi O, Fukuhara H, Kurokawa M, Izumi T, Suzuki H, Naito S, et al. Reinforcement of the Triple D score with simple addition of the intrarenal location for the prediction of the stone-free rate after shock wave lithotripsy for renal stones 10–20 mm in diameter. Int Urol Nephrol . 2019;51:239–45. doi.org/10.1007/s11255-018-02066-1 Keskin SK, Spencer M, Lovegrove C, Turney BW. The New Lithotripsy Index predicts success of shock wave lithotripsy. World J Urol . 2022;40:3049–53. doi.org/10.1007/s00345-022-04215-9 Bulut M, Dinçer E, Coşkun A, Can U, Telli O. Is Triple D score effective to predict the stone-free rate after shockwave lithotripsy in pediatric population? J Endourol . 2023;37:207–11. doi.org/10.1089/end.2022.0349 Assimos DG. Re: EAU guidelines on interventional treatment for urolithiasis. J Urol. 2016;195:659. doi.org/10.1016/j.juro.2015.12.022 Preminger GM, Tiselius HG, Assimos DG, Alken P, Buck AC, Gallucci M, et al. 2007 Guideline for the management of ureteral calculi. Eur Urol. 2007;52:1610–31. doi.org/10.1016/j.eururo.2007.09.039 Abdel-Khalek M, Sheir K, Elsobky E, Showkey S, Kenawy M. Prognostic factors for extracorporeal shock-wave lithotripsy of ureteric stones: a multivariate analysis study. Scand J Urol Nephrol . 2003;37:413–8. doi.org/10.1080/00365590310006255 Gomha MA, Sheir KZ, Showky S, Abdel-Khalek M, Mokhtar AA, Madbouly K. Can we improve the prediction of stone-free status after extracorporeal shock wave lithotripsy for ureteral stones? A neural network or a statistical model? J Urol . 2004;172:175–9. doi.org/10.1097/01.ju.0000128646.20349.27 Weld KJ, Montiglio C, Morris MS, Bush AC, Cespedes RD. Shock wave lithotripsy success for renal stones based on patient and stone computed tomography characteristics. Urology . 2007;70:1043–6; discussion 1046. doi.org/10.1016/j.urology.2007.07.074 Lee JY, Kim JH, Kang DH, et al. Stone heterogeneity index as the standard deviation of Hounsfield units: a novel predictor for shock-wave lithotripsy outcomes in ureter calculi. Sci Rep . 2016;6:23988. doi.org/10.1038/srep23988 Abdel-Khalek M, Sheir KZ, Mokhtar AA, Eraky I, Kenawy M, Bazeed M. Prediction of success rate after extracorporeal shock-wave lithotripsy of renal stones: a multivariate analysis model. Scand J Urol Nephrol . 2004;38:161–7. doi.org/10.1080/00365590310022626 Al-Ansari A, As-Sadiq K, Al-Said S, Younis N, Jaleel OA, Shokeir AA. Prognostic factors of success of extracorporeal shock wave lithotripsy (ESWL) in the treatment of renal stones. Int Urol Nephrol . 2006;38:63–7. doi.org/10.1007/s11255-005-3155-z Tables Tables 1 to 5 are available in the Supplementary Files section. Additional Declarations No competing interests reported. Supplementary Files 31.Tables.docx Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {\"props\":{\"pageProps\":{\"initialData\":{\"identity\":\"rs-6979264\",\"acceptedTermsAndConditions\":true,\"allowDirectSubmit\":true,\"archivedVersions\":[],\"articleType\":\"Research Article\",\"associatedPublications\":[],\"authors\":[{\"id\":512119609,\"identity\":\"9432817a-2eab-4596-87a9-3cef2d39bece\",\"order_by\":0,\"name\":\"Hyangmi Kim\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"Hanyang University\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Hyangmi\",\"middleName\":\"\",\"lastName\":\"Kim\",\"suffix\":\"\"},{\"id\":512119610,\"identity\":\"13ab773f-f6d3-431b-9b2f-58d37ecbba3a\",\"order_by\":1,\"name\":\"Mijeong Kim\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"Hanyang University\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Mijeong\",\"middleName\":\"\",\"lastName\":\"Kim\",\"suffix\":\"\"},{\"id\":512119611,\"identity\":\"1742cf47-402e-4ee6-8fb6-c81abb95cc2c\",\"order_by\":2,\"name\":\"Yechan Goo\",\"email\":\"\",\"orcid\":\"\",\"institution\":\"Hanyang University\",\"correspondingAuthor\":false,\"prefix\":\"\",\"firstName\":\"Yechan\",\"middleName\":\"\",\"lastName\":\"Goo\",\"suffix\":\"\"},{\"id\":512119612,\"identity\":\"43804c55-ca91-431e-a5c6-2f2814209df2\",\"order_by\":3,\"name\":\"Jung Ki Jo\",\"email\":\"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAABDElEQVRIiWNgGAWjYDACZjDJxiDP3nwMWTwBnxbGBpAWw55jaURqYQBrAYIbOWbEaTE4zvz8wcc9fHKMPWe+Pfi547C8Of8Cxg8/GNLycWo5zGbYOOMZmzE7e+92w94zhw13znjALNnDkGPZgEOLZDMPYzPPAbbExp6z2yR42w4zbrhxgEGagaHCAJctcC0NN3KeSf5tO2wP1ML8G58WfmaEFjZpoC2JG843sAFtycGjhc1w5owDbMbAQDaTlm1LT95wg7HNsscgDacWNv7DDz58OHBMDhiVzyTftlnbbjh/+PCNHxXJOLVAATzimxkYJBIbgCFJQAMDQw2MUQd06wGCykfBKBgFo2BkAQDd+1map1vXtAAAAABJRU5ErkJggg==\",\"orcid\":\"\",\"institution\":\"Hanyang University Medical Center\",\"correspondingAuthor\":true,\"prefix\":\"\",\"firstName\":\"Jung\",\"middleName\":\"Ki\",\"lastName\":\"Jo\",\"suffix\":\"\"}],\"badges\":[],\"createdAt\":\"2025-06-26 03:53:35\",\"currentVersionCode\":1,\"declarations\":\"\",\"doi\":\"10.21203/rs.3.rs-6979264/v1\",\"doiUrl\":\"https://doi.org/10.21203/rs.3.rs-6979264/v1\",\"draftVersion\":[],\"editorialEvents\":[],\"editorialNote\":\"\",\"failedWorkflow\":false,\"files\":[{\"id\":91505183,\"identity\":\"a94d14c1-3050-4780-a2ef-6b3009ad722d\",\"added_by\":\"auto\",\"created_at\":\"2025-09-17 08:14:58\",\"extension\":\"png\",\"order_by\":1,\"title\":\"Figure 1\",\"display\":\"\",\"copyAsset\":false,\"role\":\"figure\",\"size\":172823,\"visible\":true,\"origin\":\"\",\"legend\":\"\\u003cp\\u003e\\u003cstrong\\u003eBoxplot comparison of input variables by ESWL outcome (success vs. failure)\\u003c/strong\\u003e\\u003c/p\\u003e\",\"description\":\"\",\"filename\":\"1.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-6979264/v1/d08bc58821c91c37c0b186a4.png\"},{\"id\":91505178,\"identity\":\"7c81fa63-f7c1-41fe-a507-368b596cf02f\",\"added_by\":\"auto\",\"created_at\":\"2025-09-17 08:14:58\",\"extension\":\"png\",\"order_by\":2,\"title\":\"Figure 2\",\"display\":\"\",\"copyAsset\":false,\"role\":\"figure\",\"size\":58552,\"visible\":true,\"origin\":\"\",\"legend\":\"\\u003cp\\u003e\\u003cstrong\\u003eKey predictors of ESWL success identified by random forest\\u003c/strong\\u003e\\u003c/p\\u003e\",\"description\":\"\",\"filename\":\"2.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-6979264/v1/3e0e7bc039925b5cc61c0be8.png\"},{\"id\":91509275,\"identity\":\"1b6ba120-6661-4592-b9b2-b10d9fec6794\",\"added_by\":\"auto\",\"created_at\":\"2025-09-17 08:39:11\",\"extension\":\"png\",\"order_by\":3,\"title\":\"Figure 3\",\"display\":\"\",\"copyAsset\":false,\"role\":\"figure\",\"size\":110123,\"visible\":true,\"origin\":\"\",\"legend\":\"\\u003cp\\u003e\\u003cstrong\\u003eDecision tree visualization for ESWL success prediction (max depth = 4)\\u003c/strong\\u003e\\u003c/p\\u003e\",\"description\":\"\",\"filename\":\"3.png\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-6979264/v1/7fcddcfb0e1bff6127e67e1c.png\"},{\"id\":103508285,\"identity\":\"12583195-1c1c-42f8-be38-12841c2bc07f\",\"added_by\":\"auto\",\"created_at\":\"2026-02-26 13:48:04\",\"extension\":\"pdf\",\"order_by\":0,\"title\":\"\",\"display\":\"\",\"copyAsset\":false,\"role\":\"manuscript-pdf\",\"size\":842164,\"visible\":true,\"origin\":\"\",\"legend\":\"\",\"description\":\"\",\"filename\":\"manuscript.pdf\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-6979264/v1/b06abe8f-f164-408a-ac23-efb852fb624e.pdf\"},{\"id\":91506058,\"identity\":\"d853460b-0647-436b-a405-4f7bed323d1f\",\"added_by\":\"auto\",\"created_at\":\"2025-09-17 08:22:58\",\"extension\":\"docx\",\"order_by\":1,\"title\":\"\",\"display\":\"\",\"copyAsset\":false,\"role\":\"supplement\",\"size\":223202,\"visible\":true,\"origin\":\"\",\"legend\":\"\",\"description\":\"\",\"filename\":\"31.Tables.docx\",\"url\":\"https://assets-eu.researchsquare.com/files/rs-6979264/v1/d80be925f223c8a88436ab2c.docx\"}],\"financialInterests\":\"No competing interests reported.\",\"formattedTitle\":\"Seeing Beyond the Scan: Predicting ESWL Success Through Quantitative CT Parameters\",\"fulltext\":[{\"header\":\"I. Introduction\",\"content\":\"\\u003cp\\u003eUrolithiasis affects approximately 9% of the U.S. population, with 600,000 new cases reported annually. [\\u003cspan citationid=\\\"CR1\\\" class=\\\"CitationRef\\\"\\u003e1\\u003c/span\\u003e] In South Korea, the number of treated cases increased from 301,000 in 2017 to 317,472 in 2022. [\\u003cspan citationid=\\\"CR2\\\" class=\\\"CitationRef\\\"\\u003e2\\u003c/span\\u003e] Clinical management depends on stone size, location, and composition, ranging from observation to pharmacological dissolution, extracorporeal shock wave lithotripsy (ESWL), or surgery. Very small stones (\\u0026le;\\u0026thinsp;4 mm) may spontaneously pass through without causing severe symptoms or renal dysfunction. As the incidence of urolithiasis continues to increase, accurate and timely treatment selection is becoming increasingly important to minimize patient burden and optimize clinical outcomes.\\u003c/p\\u003e\\u003cp\\u003eESWL is a noninvasive procedure using external shock waves, preferred for treating small stones that are unlikely to pass naturally. It offers minimal pain, outpatient treatment, rapid recovery, and high success rates (80\\u0026ndash;90%). Treatment success is influenced by factors related to the stone (e.g., size, density), the patient (e.g., sex, age, and renal function), and ESWL treatment itself (e.g., energy escalation). [\\u003cspan citationid=\\\"CR3\\\" class=\\\"CitationRef\\\"\\u003e3\\u003c/span\\u003e] Although prior studies have explored factors, most have focused on identifying significant variables without providing guidance on prioritization or proposing standardized cut-off values to aid in patient selection. Moreover, traditional statistical methods often fail to fully capture the interactions between these variables, limiting their clinical applicability.\\u003c/p\\u003e\\u003cp\\u003eTo address these gaps, this study aims to systematically rank the key predictors of ESWL success using machine learning models and propose clinically meaningful cut-off values supporting stepwise, evidence-based decisions. By bridging predictive modeling and clinical utility, our approach seeks to provide a practical algorithm for clinicians when determining ESWL suitability.\\u003c/p\\u003e\"},{\"header\":\"II. Methods\",\"content\":\"\\u003cp\\u003eThis study was approved by the Institutional Review Board of H Hospital, South Korea (IRB No. 2022-08-017). We retrospectively analyzed the data of 488 patients who visited H Hospital for urolithiasis between 2011 and 2020. The dataset included blood and urine test results, computed tomography (CT) imaging findings, and treatment outcomes following ESWL. To predict ESWL success rates, we applied logistic regression, random forest, and decision tree models based on stone-related factors (volume, location, Hounsfield unit (HU), stone heterogeneity index (SHI), skin-to-stone distance (SSD), triple D, and quadruple D), patient characteristics (age and sex), and treatment-related variables (shock intensity and delivery efficiency).\\u003c/p\\u003e\\u003cp\\u003e\\u003cb\\u003eStone Volume.\\u003c/b\\u003e Smaller stones are consistently associated with higher stone-free rates following a single ESWL session and are recognized as independent predictors of treatment success. [\\u003cspan additionalcitationids=\\\"CR4\\\" citationid=\\\"CR3\\\" class=\\\"CitationRef\\\"\\u003e3\\u003c/span\\u003e\\u0026ndash;\\u003cspan citationid=\\\"CR5\\\" class=\\\"CitationRef\\\"\\u003e5\\u003c/span\\u003e] The European Association of Urology (EAU) Guidelines (2023) recommend ESWL as first-line therapy for stones\\u0026thinsp;\\u0026le;\\u0026thinsp;20 mm, with optimal outcomes for stones\\u0026thinsp;\\u0026le;\\u0026thinsp;10 mm. Prior studies reported success rates of approximately 85% for stone volumes\\u0026thinsp;\\u0026le;\\u0026thinsp;150 mm\\u0026sup3;. [\\u003cspan citationid=\\\"CR3\\\" class=\\\"CitationRef\\\"\\u003e3\\u003c/span\\u003e]\\u003c/p\\u003e\\u003cp\\u003e\\u003cb\\u003eStone Location.\\u003c/b\\u003e Stones located in the upper and mid-renal calyces tend to respond better to ESWL, whereas lower pole stones are less likely to be successfully cleared. ESWL is generally recommended for proximal ureteral stones\\u0026thinsp;\\u0026lt;\\u0026thinsp;10 mm, with reported success rates around 70%. [\\u003cspan citationid=\\\"CR4\\\" class=\\\"CitationRef\\\"\\u003e4\\u003c/span\\u003e\\u0026ndash;\\u003cspan citationid=\\\"CR5\\\" class=\\\"CitationRef\\\"\\u003e5\\u003c/span\\u003e]\\u003c/p\\u003e\\u003cp\\u003e\\u003cb\\u003eHU.\\u003c/b\\u003e Stone fragility is correlated with density. Hard stones, such as cystine or calcium oxalate monohydrate, showed higher HU and lower ESWL success rates. [\\u003cspan citationid=\\\"CR6\\\" class=\\\"CitationRef\\\"\\u003e6\\u003c/span\\u003e] Stones with HU\\u0026thinsp;\\u0026gt;\\u0026thinsp;1000 are associated with higher failure risks, whereas those with HU\\u0026thinsp;\\u0026le;\\u0026thinsp;1000 demonstrate significantly higher success. [\\u003cspan citationid=\\\"CR4\\\" class=\\\"CitationRef\\\"\\u003e4\\u003c/span\\u003e, \\u003cspan additionalcitationids=\\\"CR8\\\" citationid=\\\"CR7\\\" class=\\\"CitationRef\\\"\\u003e7\\u003c/span\\u003e\\u0026ndash;\\u003cspan citationid=\\\"CR9\\\" class=\\\"CitationRef\\\"\\u003e9\\u003c/span\\u003e]\\u003c/p\\u003e\\u003cp\\u003e\\u003cb\\u003eSHI.\\u003c/b\\u003e The SHI quantifies the internal variability of stone. A higher SHI indicates greater structural heterogeneity, facilitating shockwave fragmentation and improving ESWL success. [\\u003cspan citationid=\\\"CR10\\\" class=\\\"CitationRef\\\"\\u003e10\\u003c/span\\u003e, \\u003cspan citationid=\\\"CR11\\\" class=\\\"CitationRef\\\"\\u003e11\\u003c/span\\u003e]\\u003c/p\\u003e\\u003cp\\u003e\\u003cb\\u003eSSD.\\u003c/b\\u003e An SSD of \\u0026le;\\u0026thinsp;10 cm has been associated with better ESWL outcomes. [\\u003cspan citationid=\\\"CR4\\\" class=\\\"CitationRef\\\"\\u003e4\\u003c/span\\u003e, \\u003cspan additionalcitationids=\\\"CR13\\\" citationid=\\\"CR12\\\" class=\\\"CitationRef\\\"\\u003e12\\u003c/span\\u003e\\u0026ndash;\\u003cspan citationid=\\\"CR14\\\" class=\\\"CitationRef\\\"\\u003e14\\u003c/span\\u003e] However, recent studies suggest that stone-related characteristics may be more predictive of treatment success than patient body habitus alone. [\\u003cspan citationid=\\\"CR15\\\" class=\\\"CitationRef\\\"\\u003e15\\u003c/span\\u003e]\\u003c/p\\u003e\\u003cp\\u003e\\u003cb\\u003eTriple D Score.\\u003c/b\\u003e The triple D Score integrates stone volume, HU, and SSD to predict the likelihood of ESWL success. [\\u003cspan citationid=\\\"CR3\\\" class=\\\"CitationRef\\\"\\u003e3\\u003c/span\\u003e, \\u003cspan citationid=\\\"CR16\\\" class=\\\"CitationRef\\\"\\u003e16\\u003c/span\\u003e] Higher scores have been correlated with improved outcomes. In this study, we adopted cut-off values based on previous research: stone volume\\u0026thinsp;\\u0026le;\\u0026thinsp;150 mm\\u0026sup3;, SSD\\u0026thinsp;\\u0026le;\\u0026thinsp;9 cm, and HU\\u0026thinsp;\\u0026le;\\u0026thinsp;1000.\\u003c/p\\u003e\\u003cp\\u003e\\u003cb\\u003eQuadruple D Score.\\u003c/b\\u003e The quadruple D score extends the triple D score by incorporating lower pole stone (LPS) status. [\\u003cspan citationid=\\\"CR17\\\" class=\\\"CitationRef\\\"\\u003e17\\u003c/span\\u003e] It uses the criteria of stone volume\\u0026thinsp;\\u0026le;\\u0026thinsp;150 mm\\u0026sup3;, SSD\\u0026thinsp;\\u0026le;\\u0026thinsp;9 cm, HU\\u0026thinsp;\\u0026le;\\u0026thinsp;1000, and absence of LPS, with higher scores being associated with improved ESWL success rates. [\\u003cspan citationid=\\\"CR11\\\" class=\\\"CitationRef\\\"\\u003e11\\u003c/span\\u003e] In the present study, the cut-off point for the quadruple D score was based on that of the triple D score, with the addition of stone location (lower).\\u003c/p\\u003e\\u003cp\\u003e\\u003cb\\u003eSex.\\u003c/b\\u003e Females may have slightly higher ESWL success rates than males [\\u003cspan citationid=\\\"CR7\\\" class=\\\"CitationRef\\\"\\u003e7\\u003c/span\\u003e]; however, sex is not considered an independent predictor when compared to factors such as stone size, location, or density. [\\u003cspan citationid=\\\"CR9\\\" class=\\\"CitationRef\\\"\\u003e9\\u003c/span\\u003e]\\u003c/p\\u003e\\u003cp\\u003e\\u003cb\\u003eAge.\\u003c/b\\u003e Younger patients tend to achieve higher ESWL success rates because of their more efficient renal and ureteral function. [\\u003cspan citationid=\\\"CR18\\\" class=\\\"CitationRef\\\"\\u003e18\\u003c/span\\u003e] In contrast, older patients may show worse outcomes owing to anatomical changes and harder stone compositions.\\u003c/p\\u003e\\u003cp\\u003e\\u003cb\\u003eShock Wave (Energy) Intensity and Efficiency.\\u003c/b\\u003e Gradual escalation of the shock wave intensity has been shown to enhance stone fragmentation while minimizing renal injury. [\\u003cspan citationid=\\\"CR5\\\" class=\\\"CitationRef\\\"\\u003e5\\u003c/span\\u003e] Recent studies have also emphasized the importance of efficiency metrics, such as the ratio of delivered energy to stone volume (SMLI/SV). [\\u003cspan citationid=\\\"CR3\\\" class=\\\"CitationRef\\\"\\u003e3\\u003c/span\\u003e]\\u003c/p\\u003e\\u003cp\\u003eWe used Python (version 3.10) and the scikit-learn library (version 1.2.2) for data preprocessing and machine learning-based predictive modeling (logistic regression, random forest, and decision tree), based on the aforementioned variables.\\u003c/p\\u003e\"},{\"header\":\"III. Results\",\"content\":\"\\u003cp\\u003eA total of 488 patients (345 males and 143 females) were included in the study, with an average age of 50 years; 73.16% were aged \\u0026gt; 40 years. In addition, there were 383 successful cases of stone removal following extracorporeal shock wave lithotripsy (ESWL) and 105 cases in which the treatment failed.\\u003c/p\\u003e\\n\\u003cp\\u003eData pre-processing involved random forest imputation for categorical variables and median replacement for continuous variables. Feature engineering created three derived variables: the total ESWL energy per stone volume, HU, and SSD. Three predictive models—logistic regression, random forest, and decision tree—were developed to compare the performance and identify key predictors.\\u003c/p\\u003e\\n\\u003cp\\u003ePrior to analysis, the distribution of eight continuous variables, excluding the derived variables, was examined in relation to ESWL outcomes (success vs. failure), as illustrated, in Figure 1. The distributions of age, total ESWL energy, stone volume, HU, SHI, and SSD appeared similar between the success and failure groups, which contrasts with findings from previous studies. However, because ESWL success may depend on subtle differences in variable values, we proceeded to identify significant predictors using logistic regression, random forest, and decision tree analyses.\\u003c/p\\u003e\\n\\u003cp\\u003e\\u003cstrong\\u003e3.1. Logistic Regression\\u003c/strong\\u003e\\u003c/p\\u003e\\n\\u003cp\\u003eLogistic regression was initially employed as a baseline model because of its ability to statistically evaluate the effect of each variable on ESWL success and maintain consistency with prior studies.\\u003c/p\\u003e\\n\\u003cp\\u003eThe model demonstrated a strong performance, achieving an accuracy of 94.3% and an area under the curve (AUC) of 0.883. Sensitivity was notably high at 98.9% (365/369 successful cases correctly classified), indicating excellent identification of likely successful candidates. However, specificity was relatively low at 44.1% (15/34 failures correctly classified), reflecting limited ability to detect treatment failures.\\u003c/p\\u003e\\n\\u003cp\\u003eThese results highlight the model’s effectiveness in reducing missed treatment opportunities, which is a key priority in clinical decision-making. The significant predictors of ESWL success included stone location in the upper ureter (odds ratio [OR] = 9.81, p \\u0026lt; 0.001), renal pelvis (OR = 6.47, p = 0.01), higher SHI (OR = 1.02, p \\u0026lt; 0.001), and lower HU (OR = 0.99, p \\u0026lt; 0.001). Notably, in contrast to previous studies, a greater SSD was positively associated with ESWL success (OR = 1.03, p = 0.01). A higher total ESWL energy per unit stone volume was associated with lower success rates, suggesting that the efficiency of energy delivery may be more important than the absolute energy applied. No significant associations were found for age, sex, stone volume, total ESWL energy, or other derived variables.\\u003c/p\\u003e\\n\\u003cp\\u003eLogistic regression analysis is valued for its interpretability and ability to support statistical inference. However, it has limitations in capturing complex nonlinear relationships between variables and is sensitive to multicollinearity, which can reduce its predictive performance compared to tree-based models. To address these limitations and enhance predictive accuracy, we utilized the random forest and decision tree methods.\\u003c/p\\u003e\\n\\u003cp\\u003e\\u003cstrong\\u003e3.2. Random Forest Model\\u003c/strong\\u003e\\u003c/p\\u003e\\n\\u003cp\\u003eAs mentioned earlier, the random forest model was implemented to capture complex nonlinear relationships, which is a limitation of logistic regression, and to improve model stability. The data were split into an 80% training dataset and a 20% test dataset, using 100 decision trees (n_estimators = 100).\\u003c/p\\u003e\\n\\u003cp\\u003eThe random forest analysis showed that the model had an accuracy of 92.6%, a sensitivity of 98.6%, a specificity of 28.6%, and an AUC of 0.93, which was better than that of the logistic regression model (AUC = 0.883). This high sensitivity was clinically meaningful, as it minimized the risk of missing patients who could be successfully treated with ESWL.\\u003c/p\\u003e\\n\\u003cp\\u003eThe results of the variable importance analysis identified stone volume, HU, and SHI as the most influential predictors of ESWL success. Figure 2 presents the relative importance of these variables as determined by the random forest model. Although these results were not completely consistent with the results of the logistic regression analysis, some of the significant variables were consistent with the results of previous studies and the logistic regression analysis. Stone volume, HU, and SHI were reliable predictors of ESWL success.\\u003c/p\\u003e\\n\\u003cp\\u003e\\u003cstrong\\u003e3.3. Decision Tree Model\\u003c/strong\\u003e\\u003c/p\\u003e\\n\\u003cp\\u003eAlthough logistic regression and random forest analyses demonstrated strong predictive performance, clearer criteria are needed for clinical decision-making by healthcare professionals. Therefore, using the same dataset, we implemented a decision tree model with a maximum depth of four to prevent overfitting and enhance visualization. The model was built based on the key variables and criteria that affect ESWL success for urolithiasis. It presented clear decision rules and classification paths, providing an intuitive understanding for clinicians.\\u003c/p\\u003e\\n\\u003cp\\u003eThe decision tree achieved an accuracy of 91.4%, a sensitivity of 95.9%, a specificity of 42.9%, and an AUC of 0.559. Although its overall performance was lower than that of logistic regression and random forest, its high sensitivity highlights its potential value in supporting clinical decisions.\\u003c/p\\u003e\\n\\u003cp\\u003eThe decision tree model offers strong explanatory power by capturing nonlinear relationships and visually representing the decision rules (Figure 3). The primary classification factor was stone volume (≤ 27.99mm³), which was significantly associated with a higher ESWL success rate, identifying smaller stones as a major predictor. Further stratification based on stone size, location, and SHI revealed that among patients with clearly defined stone locations, those with SHI ≥ 50 had a higher success rate, suggesting that the greater internal heterogeneity leads to better fragmentation. For stones with volumes ≥ 27.99 mm³, HU became the next classification criterion. Stones with HU ≥ 393.61 (mean HU 486.66 in the success group) and SHI ≤ 256.18 had a higher failure rate, indicating that more homogeneous stone material, is harder to fragment with ESWL. These findings highlight the utility of decision trees as interpretable tools that provide intuitive predictive paths to guide clinical decision-making for ESWL.\\u003c/p\\u003e\\n\\u003cp\\u003eIn summary, we employed three machine learning approaches (logistic regression, random forest, and decision trees) to predict ESWL success based on features obtained from NCCT, patient demographic characteristics, and treatment variables. Model performance was evaluated using standard classification metrics: accuracy, sensitivity, specificity, and AUC. The use of three approaches was intentional to leverage the strengths of each method. Logistic regression was chosen to allow comparison with previous studies through statistical inference, and is served as a baseline model for the random forest and decision tree models. The random forest model facilitated robust modeling of nonlinear relationships, while the decision tree provided an interpretable, rule-based decision path that enhances clinical applicability. This comparative approach, using multiple models, was designed to strengthen predictive validity and ensure practical relevance for real-world clinical decision-making.\\u003c/p\\u003e\"},{\"header\":\"IV. Discussion\",\"content\":\"\\u003cp\\u003eESWL is a noninvasive procedure that uses externally generated shock waves to fragment urinary stones. It is widely used for treating renal and ureteral stones, particularly small ones, without requiring hospitalization, anesthesia, or surgery. ESWL remains an effective first-line treatment for appropriately selected patients. [\\u003cspan citationid=\\\"CR3\\\" class=\\\"CitationRef\\\"\\u003e3\\u003c/span\\u003e, \\u003cspan citationid=\\\"CR9\\\" class=\\\"CitationRef\\\"\\u003e9\\u003c/span\\u003e, \\u003cspan citationid=\\\"CR19\\\" class=\\\"CitationRef\\\"\\u003e19\\u003c/span\\u003e\\u0026ndash;\\u003cspan citationid=\\\"CR20\\\" class=\\\"CitationRef\\\"\\u003e20\\u003c/span\\u003e] For small stones, repeated ESWL can achieve stone-free rates of up to 90%. [\\u003cspan citationid=\\\"CR21\\\" class=\\\"CitationRef\\\"\\u003e21\\u003c/span\\u003e] Identifying patients most likely to benefit is clinically important. Various predictive models have evaluated factors associated with success. Prior studies consistently show that smaller stone size, [\\u003cspan citationid=\\\"CR3\\\" class=\\\"CitationRef\\\"\\u003e3\\u003c/span\\u003e, \\u003cspan citationid=\\\"CR4\\\" class=\\\"CitationRef\\\"\\u003e4\\u003c/span\\u003e, \\u003cspan citationid=\\\"CR8\\\" class=\\\"CitationRef\\\"\\u003e8\\u003c/span\\u003e, \\u003cspan citationid=\\\"CR13\\\" class=\\\"CitationRef\\\"\\u003e13\\u003c/span\\u003e, \\u003cspan additionalcitationids=\\\"CR23\\\" citationid=\\\"CR22\\\" class=\\\"CitationRef\\\"\\u003e22\\u003c/span\\u003e\\u0026ndash;\\u003cspan citationid=\\\"CR24\\\" class=\\\"CitationRef\\\"\\u003e24\\u003c/span\\u003e] lower HU values, [\\u003cspan citationid=\\\"CR6\\\" class=\\\"CitationRef\\\"\\u003e6\\u003c/span\\u003e] greater heterogeneity, [\\u003cspan citationid=\\\"CR25\\\" class=\\\"CitationRef\\\"\\u003e25\\u003c/span\\u003e] and favorable locations (renal pelvis or upper ureter) [\\u003cspan citationid=\\\"CR7\\\" class=\\\"CitationRef\\\"\\u003e7\\u003c/span\\u003e, \\u003cspan additionalcitationids=\\\"CR23\\\" citationid=\\\"CR22\\\" class=\\\"CitationRef\\\"\\u003e22\\u003c/span\\u003e\\u0026ndash;\\u003cspan citationid=\\\"CR24\\\" class=\\\"CitationRef\\\"\\u003e24\\u003c/span\\u003e, \\u003cspan citationid=\\\"CR26\\\" class=\\\"CitationRef\\\"\\u003e26\\u003c/span\\u003e\\u0026ndash;\\u003cspan citationid=\\\"CR27\\\" class=\\\"CitationRef\\\"\\u003e27\\u003c/span\\u003e] predict higher success. However traditional statistical methods may not fully capture complex interactions between variables. [\\u003cspan citationid=\\\"CR3\\\" class=\\\"CitationRef\\\"\\u003e3\\u003c/span\\u003e, \\u003cspan citationid=\\\"CR13\\\" class=\\\"CitationRef\\\"\\u003e13\\u003c/span\\u003e, \\u003cspan citationid=\\\"CR22\\\" class=\\\"CitationRef\\\"\\u003e22\\u003c/span\\u003e, \\u003cspan citationid=\\\"CR26\\\" class=\\\"CitationRef\\\"\\u003e26\\u003c/span\\u003e] To address this, we applied three machine learning models\\u0026mdash;logistic regression, random forest, and decision tree\\u0026mdash;to predict ESWL outcomes.\\u003c/p\\u003e\\u003cp\\u003eConsistent with prior studies, this study confirmed that a small stone volume (\\u0026le;\\u0026thinsp;27.99 mm\\u0026sup3;), HU (\\u0026lt;\\u0026thinsp;393.61), SHI (\\u0026ge;\\u0026thinsp;50), and stone location (renal pelvis or upper ureter) were key predictors of ESWL success in order of importance. Notably, SSD showed a positive correlation with ESWL success, contrasting with prior studies, but this alone may not reliably predict outcomes due to potential confounding factors in the logistic regression model. Although both the logistic regression and random forest models performed well, they lacked explainability for clinical decision-making. Therefore, we developed a decision tree model to visualize the classification process and threshold for key predictors. Based on thresholds from our models, we propose a stepwise clinical decision-making algorithm for ESWL candidacy. There thresholds can be assessed on routine NCCT scans, enabling clinician to stratify patients and select appropriate treatment. Integrating these quantitative criteria into the pre-treatment evaluation can enhance ESWL decision-making and outcomes.\\u003c/p\\u003e\\u003cp\\u003eThis study has several limitations. First, the single-center retrospective design may limit generalizability, underscoring the need for validation in external cohorts. Second, the small number of failure cases may have reduced the model\\u0026rsquo;s specificity; therefore, future studies should consider using oversampling or data balancing techniques to address this issue. Third, although AUC comparisons were conducted, testing (e.g., the DeLong test) should be applied in future studies to confirm significant differences in model performance and strengthen the robustness of validation.\\u003c/p\\u003e\\u003cp\\u003eFurther research should validate the findings of this study in a larger patient population to enhance the generalizability of the results. Additionally, studies should investigate how to integrate this system into real-world clinical workflows to support real-time decision-making.\\u003c/p\\u003e\"},{\"header\":\"V. Conclusion\",\"content\":\"\\u003cp\\u003eIn this study, we evaluated three machine learning models to predict the success rate of ESWL and found that a small stone volume, small HU, large SHI, and stone location (renal pelvis or upper ureter) were the key predictors. As these features can be easily obtained from NCCT, they serve as practical indicators to help clinicians make timely and effective decisions regarding ESWL procedures, thereby supporting more efficient clinical management.\\u003c/p\\u003e\"},{\"header\":\"Declarations\",\"content\":\"\\u003cp\\u003e\\u003cstrong\\u003eEthics approval and consent to participate\\u003c/strong\\u003e\\u003c/p\\u003e\\n\\u003cp\\u003eThis retrospective study was approved by the Institutional Review Board of Hanyang University Medical Center (IRB No. 2022-08-017). The requirement for informed consent was waived due to the retrospective nature of the study.\\u0026nbsp;\\u003c/p\\u003e\\n\\u003cp\\u003e\\u003cstrong\\u003eConsent for publication\\u003c/strong\\u003e\\u003c/p\\u003e\\n\\u003cp\\u003eNot applicable\\u0026nbsp;\\u003c/p\\u003e\\n\\u003cp\\u003e\\u003cstrong\\u003eFunding\\u003c/strong\\u003e\\u003c/p\\u003e\\n\\u003cp\\u003eNo funding was received for conducting this study.\\u0026nbsp;\\u003c/p\\u003e\\n\\u003cp\\u003e\\u003cstrong\\u003eCompeting interests\\u003c/strong\\u003e\\u003c/p\\u003e\\n\\u003cp\\u003eThe authors declare that they have no competing interests.\\u0026nbsp;\\u003c/p\\u003e\\n\\u003cp\\u003e\\u003cstrong\\u003eAuthor Contribution\\u003c/strong\\u003e\\u003c/p\\u003e\\n\\u003cp\\u003eH.M.K. designed the study, conducted data analysis, and drafted the manuscript. Y.C.G. collected the data. M.J.K. contributed to data collection and manuscript editing. J.K.J. supervised the entire research process as the senior advisor and provided overall guidance throughout the study. All authors reviewed and approved the final version of the manuscript.\\u003c/p\\u003e\"},{\"header\":\"References\",\"content\":\"\\u003col\\u003e\\n\\u003cli\\u003eUreteroscope mark size share COVID. \\u003cem\\u003eFortune Business Insights\\u003c/em\\u003e. Published May 19, 2025. https://www.fortunebusinessinsights.com/ureteroscope-market-103455. Accessed March 3, 2026.\\u003c/li\\u003e\\n\\u003cli\\u003eHealth Insurance Review \\u0026amp; Assessment Service. Report on statistics of diseases and medical procedures in daily life. 2023.\\u003c/li\\u003e\\n\\u003cli\\u003eSnicorius M, Drevinskaitė M, Miglinas M, et al. A prospective study on the impact of clinical factors and adjusted Triple D system for success rate of ESWL. \\u003cem\\u003eMedicina (Kaunas)\\u003c/em\\u003e. 2023;59:1827. doi.org/10.3390/medicina59101827\\u003c/li\\u003e\\n\\u003cli\\u003eBajaj M, Smith R, Rice M, Zargar-Shoshtari K. Predictors of success following extracorporeal shock-wave lithotripsy in a contemporary cohort. \\u003cem\\u003eUrol Ann\\u003c/em\\u003e. 2021;13:282\\u0026ndash;7. doi.org/10.4103/UA.UA_155_19\\u003c/li\\u003e\\n\\u003cli\\u003eOrdon M, Andonian S, Blew B, Schuler T, Chew B, Pace KT. CUA guideline. CUA Guideline: management of ureteral calculi. \\u003cem\\u003eCan Urol Assoc J\\u003c/em\\u003e. 2015;9:E837\\u0026ndash;51. doi.org/10.5489/cuaj.3483\\u003c/li\\u003e\\n\\u003cli\\u003eYoon JH, Park S, Kim SC, Park S, Moon KH, Cheon SH, et al. Outcomes of extracorporeal shock wave lithotripsy for ureteral stones according to ESWL intensity. \\u003cem\\u003eTransl Androl Urol.\\u003c/em\\u003e 2021;10:1588\\u0026ndash;95. doi.org/10.21037/tau-20-1397\\u003c/li\\u003e\\n\\u003cli\\u003eEl-Nahas AR, El-Assmy AM, Mansour O, Sheir KZ. A prospective multivariate analysis of factors predicting stone disintegration by extracorporeal shock wave lithotripsy: the value of high-resolution noncontrast computed tomography. \\u003cem\\u003eEur Urol\\u003c/em\\u003e. 2007;51:1688\\u0026ndash;93; discussion 1693. doi.org/10.1016/j.eururo.2006.11.048\\u003c/li\\u003e\\n\\u003cli\\u003eOuzaid I, Al-Qahtani S, Dominique S, Hupertan V, Fernandez P, Hermieu JF, et al. A 970 Hounsfield units (HU) threshold of kidney stone density on non-contrast computed tomography (NCCT) improves patients\\u0026rsquo; selection for extracorporeal shockwave lithotripsy (ESWL): evidence from a prospective study. \\u003cem\\u003eBJU Int\\u003c/em\\u003e. 2012;110;110:E438\\u0026ndash;42. doi.org/10.1111/j.1464-410X.2012.10964.x\\u003c/li\\u003e\\n\\u003cli\\u003eT\\u0026uuml;rk C, Petř\\u0026iacute;k A, Sarica K, et al. EAU guidelines on interventional treatment for urolithiasis. \\u003cem\\u003eEur Urol\\u003c/em\\u003e. 2016;69:475\\u0026ndash;82. doi.org/10.1016/j.eururo.2015.07.041\\u003c/li\\u003e\\n\\u003cli\\u003eIqbal N, Hasan A, Nazar A, et al. Role of stone heterogeneity index in determining success of shock wave lithotripsy in urinary calculi. \\u003cem\\u003eJ Clin Transl Res\\u003c/em\\u003e. 2021;7:241\\u0026ndash;7.\\u003c/li\\u003e\\n\\u003cli\\u003eŞendoğan F, Bulut M, \\u0026Ccedil;anakcı C, et al. Quadruple-D score in the success rate of extracorporeal shock wave lithotripsy of renal stones in pediatric population. \\u003cem\\u003eUrolithiasis\\u003c/em\\u003e. 2024;52:163. doi.org/10.1007/s00240-024-01657-1\\u003c/li\\u003e\\n\\u003cli\\u003eCho KS, Jung HD, Ham WS, et al. Optimal skin-to-stone distance is a positive predictor for successful outcomes in upper ureter calculi following extracorporeal shock wave lithotripsy: a Bayesian model averaging approach. \\u003cem\\u003ePLOS One\\u003c/em\\u003e. 2015;10:e0144912. doi.org/10.1371/journal.pone.0144912\\u003c/li\\u003e\\n\\u003cli\\u003ePareek G, Armenakas NA, Panagopoulos G, Bruno JJ, Fracchia JA. Extracorporeal shock wave lithotripsy success based on body mass index and Hounsfield units. \\u003cem\\u003eUrology.\\u003c/em\\u003e 2005;65:33\\u0026ndash;6. doi.org/10.1016/j.urology.2004.08.004\\u003c/li\\u003e\\n\\u003cli\\u003ePatel T, Kozakowski K, Hruby G, Gupta M. Skin to stone distance is an independent predictor of stone-free status following shock wave lithotripsy. \\u003cem\\u003eJ Endourol\\u003c/em\\u003e. 2009;23:1383\\u0026ndash;5. doi.org/10.1089/end.2009.0394.\\u003c/li\\u003e\\n\\u003cli\\u003e\\u0026Ccedil;anakcı C, Din\\u0026ccedil;er E, Şimşek B, et al. Effect of tissue densities at the skin-to-stone distance on the success of shockwave lithotripsy. \\u003cem\\u003eJ Urol Surg\\u003c/em\\u003e. 2024;11:14\\u0026ndash;8. doi.org/10.4274/jus.galenos.2023.2023.0010\\u003c/li\\u003e\\n\\u003cli\\u003eTran TY, McGillen K, Cone EB, Pareek G. Triple D score is a reportable predictor of shockwave lithotripsy stone-free rates. \\u003cem\\u003eJ Endourol\\u003c/em\\u003e. 2015;29:226\\u0026ndash;30. doi.org/10.1089/end.2014.0212\\u003c/li\\u003e\\n\\u003cli\\u003eIchiyanagi O, Fukuhara H, Kurokawa M, Izumi T, Suzuki H, Naito S, et al. Reinforcement of the Triple D score with simple addition of the intrarenal location for the prediction of the stone-free rate after shock wave lithotripsy for renal stones 10\\u0026ndash;20 mm in diameter. \\u003cem\\u003eInt Urol Nephrol\\u003c/em\\u003e. 2019;51:239\\u0026ndash;45. doi.org/10.1007/s11255-018-02066-1\\u003c/li\\u003e\\n\\u003cli\\u003eKeskin SK, Spencer M, Lovegrove C, Turney BW. The New Lithotripsy Index predicts success of shock wave lithotripsy. \\u003cem\\u003eWorld J Urol\\u003c/em\\u003e. 2022;40:3049\\u0026ndash;53. doi.org/10.1007/s00345-022-04215-9\\u003c/li\\u003e\\n\\u003cli\\u003eBulut M, Din\\u0026ccedil;er E, Coşkun A, Can U, Telli O. Is Triple D score effective to predict the stone-free rate after shockwave lithotripsy in pediatric population? \\u003cem\\u003eJ Endourol\\u003c/em\\u003e. 2023;37:207\\u0026ndash;11. doi.org/10.1089/end.2022.0349\\u003c/li\\u003e\\n\\u003cli\\u003eAssimos DG. Re: EAU guidelines on interventional treatment for urolithiasis. \\u003cem\\u003eJ Urol.\\u003c/em\\u003e 2016;195:659. doi.org/10.1016/j.juro.2015.12.022\\u003c/li\\u003e\\n\\u003cli\\u003ePreminger GM, Tiselius HG, Assimos DG, Alken P, Buck AC, Gallucci M, et al. 2007 Guideline for the management of ureteral calculi. \\u003cem\\u003eEur Urol.\\u003c/em\\u003e 2007;52:1610\\u0026ndash;31. doi.org/10.1016/j.eururo.2007.09.039\\u003c/li\\u003e\\n\\u003cli\\u003eAbdel-Khalek M, Sheir K, Elsobky E, Showkey S, Kenawy M. Prognostic factors for extracorporeal shock-wave lithotripsy of ureteric stones: a multivariate analysis study. \\u003cem\\u003eScand J Urol Nephrol\\u003c/em\\u003e. 2003;37:413\\u0026ndash;8. doi.org/10.1080/00365590310006255\\u003c/li\\u003e\\n\\u003cli\\u003eGomha MA, Sheir KZ, Showky S, Abdel-Khalek M, Mokhtar AA, Madbouly K. Can we improve the prediction of stone-free status after extracorporeal shock wave lithotripsy for ureteral stones? A neural network or a statistical model? \\u003cem\\u003eJ Urol\\u003c/em\\u003e. 2004;172:175\\u0026ndash;9. doi.org/10.1097/01.ju.0000128646.20349.27\\u003c/li\\u003e\\n\\u003cli\\u003eWeld KJ, Montiglio C, Morris MS, Bush AC, Cespedes RD. Shock wave lithotripsy success for renal stones based on patient and stone computed tomography characteristics. \\u003cem\\u003eUrology\\u003c/em\\u003e. 2007;70:1043\\u0026ndash;6; discussion 1046. doi.org/10.1016/j.urology.2007.07.074\\u003c/li\\u003e\\n\\u003cli\\u003eLee JY, Kim JH, Kang DH, et al. Stone heterogeneity index as the standard deviation of Hounsfield units: a novel predictor for shock-wave lithotripsy outcomes in ureter calculi. \\u003cem\\u003eSci Rep\\u003c/em\\u003e. 2016;6:23988. doi.org/10.1038/srep23988\\u003c/li\\u003e\\n\\u003cli\\u003eAbdel-Khalek M, Sheir KZ, Mokhtar AA, Eraky I, Kenawy M, Bazeed M. Prediction of success rate after extracorporeal shock-wave lithotripsy of renal stones: a multivariate analysis model. \\u003cem\\u003eScand J Urol Nephrol\\u003c/em\\u003e. 2004;38:161\\u0026ndash;7. doi.org/10.1080/00365590310022626\\u003c/li\\u003e\\n\\u003cli\\u003eAl-Ansari A, As-Sadiq K, Al-Said S, Younis N, Jaleel OA, Shokeir AA. Prognostic factors of success of extracorporeal shock wave lithotripsy (ESWL) in the treatment of renal stones. \\u003cem\\u003eInt Urol Nephrol\\u003c/em\\u003e. 2006;38:63\\u0026ndash;7. doi.org/10.1007/s11255-005-3155-z\\u003c/li\\u003e\\n\\u003c/ol\\u003e\"},{\"header\":\"Tables\",\"content\":\"\\u003cp\\u003eTables 1 to 5 are available in the Supplementary Files section.\\u003c/p\\u003e\"}],\"fulltextSource\":\"\",\"fullText\":\"\",\"funders\":[],\"hasAdminPriorityOnWorkflow\":false,\"hasManuscriptDocX\":true,\"hasOptedInToPreprint\":true,\"hasPassedJournalQc\":\"\",\"hasAnyPriority\":false,\"hideJournal\":true,\"highlight\":\"\",\"institution\":\"\",\"isAcceptedByJournal\":false,\"isAuthorSuppliedPdf\":false,\"isDeskRejected\":\"\",\"isHiddenFromSearch\":false,\"isInQc\":false,\"isInWorkflow\":false,\"isPdf\":false,\"isPdfUpToDate\":true,\"isWithdrawnOrRetracted\":false,\"journal\":{\"display\":true,\"email\":\"info@researchsquare.com\",\"identity\":\"researchsquare\",\"isNatureJournal\":false,\"hasQc\":true,\"allowDirectSubmit\":true,\"externalIdentity\":\"\",\"sideBox\":\"\",\"snPcode\":\"\",\"submissionUrl\":\"/submission\",\"title\":\"Research Square\",\"twitterHandle\":\"researchsquare\",\"acdcEnabled\":true,\"dfaEnabled\":false,\"editorialSystem\":\"\",\"reportingPortfolio\":\"\",\"inReviewEnabled\":false,\"inReviewRevisionsEnabled\":true},\"keywords\":\"Clinical decision support, Extracorporeal shock wave lithotripsy (ESWL), Machine learning, Non-contrast computed tomography (NCCT), Urolithiasis\",\"lastPublishedDoi\":\"10.21203/rs.3.rs-6979264/v1\",\"lastPublishedDoiUrl\":\"https://doi.org/10.21203/rs.3.rs-6979264/v1\",\"license\":{\"name\":\"CC BY 4.0\",\"url\":\"https://creativecommons.org/licenses/by/4.0/\"},\"manuscriptAbstract\":\"\\u003cp\\u003e\\u003cb\\u003ePurpose\\u003c/b\\u003e\\u003c/p\\u003e\\u003cp\\u003eExtracorporeal shock wave lithotripsy (ESWL) is a non-invasive treatment for urolithiasis; however, predicting its success remains challenging. This study aimed to provide guidance on the key predictors and clinical criteria for anticipating the success of ESWL.\\u003c/p\\u003e\\u003cp\\u003e\\u003cb\\u003eMaterials and Methods\\u003c/b\\u003e\\u003c/p\\u003e\\u003cp\\u003eWe retrospectively analyzed 488 patients who underwent ESWL between 2011 and 2020. Predictive features included patient demographics, stone characteristics on non-contrast computed tomography (NCCT), composite scores (triple D, quadruple D), and ESWL energy parameters. Logistic regression, random forest, and decision tree models were used to identify key predictors of treatment success based on accuracy, sensitivity, and specificity.\\u003c/p\\u003e\\u003cp\\u003e\\u003cb\\u003eResults\\u003c/b\\u003e\\u003c/p\\u003e\\u003cp\\u003eLogistic regression showed 94.3% accuracy (area under the curve [AUC] 0.883), with 92.6% (AUC 0.936) for random forest and 91.4% (AUC 0.559) for decision tree. Small stones, lower Hounsfield unit, higher stone heterogeneity index, and favorable locations (pelvis/upper) were consistent predictors of success. Although the sample size was robust, the single-center design may have limited the external validity of the findings.\\u003c/p\\u003e\\u003cp\\u003e\\u003cb\\u003eConclusions\\u003c/b\\u003e\\u003c/p\\u003e\\u003cp\\u003eNCCT-derived stone characteristics are sufficient to predict ESWL outcomes. This study provides a threshold-based guidance for each predictor, enabling clinicians to make timely and informed decisions regarding ESWL in patients with urolithiasis.\\u003c/p\\u003e\",\"manuscriptTitle\":\"Seeing Beyond the Scan: Predicting ESWL Success Through Quantitative CT Parameters\",\"msid\":\"\",\"msnumber\":\"\",\"nonDraftVersions\":[{\"code\":1,\"date\":\"2025-09-17 08:14:54\",\"doi\":\"10.21203/rs.3.rs-6979264/v1\",\"editorialEvents\":[{\"type\":\"communityComments\",\"content\":0}],\"status\":\"published\",\"journal\":{\"display\":true,\"email\":\"info@researchsquare.com\",\"identity\":\"researchsquare\",\"isNatureJournal\":false,\"hasQc\":true,\"allowDirectSubmit\":true,\"externalIdentity\":\"\",\"sideBox\":\"\",\"snPcode\":\"\",\"submissionUrl\":\"/submission\",\"title\":\"Research Square\",\"twitterHandle\":\"researchsquare\",\"acdcEnabled\":true,\"dfaEnabled\":false,\"editorialSystem\":\"\",\"reportingPortfolio\":\"\",\"inReviewEnabled\":false,\"inReviewRevisionsEnabled\":true}}],\"origin\":\"\",\"ownerIdentity\":\"a9327967-1429-4e7b-b9dc-635b87a60fd6\",\"owner\":[],\"postedDate\":\"September 17th, 2025\",\"published\":true,\"recentEditorialEvents\":[],\"rejectedJournal\":[],\"revision\":\"\",\"amendment\":\"\",\"status\":\"posted\",\"subjectAreas\":[],\"tags\":[],\"updatedAt\":\"2026-02-26T04:24:39+00:00\",\"versionOfRecord\":[],\"versionCreatedAt\":\"2025-09-17 08:14:54\",\"video\":\"\",\"vorDoi\":\"\",\"vorDoiUrl\":\"\",\"workflowStages\":[]},\"version\":\"v1\",\"identity\":\"rs-6979264\",\"journalConfig\":\"researchsquare\"},\"__N_SSP\":true},\"page\":\"/article/[identity]/[[...version]]\",\"query\":{\"redirect\":\"/article/rs-6979264\",\"identity\":\"rs-6979264\",\"version\":[\"v1\"]},\"buildId\":\"8U1c8b4HqxoKbykW_rLl7\",\"isFallback\":false,\"isExperimentalCompile\":false,\"dynamicIds\":[84888],\"gssp\":true,\"scriptLoader\":[]}","source_license":"CC-BY-4.0","license_restricted":false}