World journal of urology · 2026
Risk factors and machine learning-based prediction of difficult ureter during retrograde ureteroscopic lithotripsy.
СтатьяНазвание и аннотация: Перевод пока не подготовлен
Обработка статьи · этапы, ошибки и управление
Загрузка состояния…
Ошибки и ограничения
История · было → стало
Состояние обновляется каждые 5 секунд, пока блок открыт. Процент показывается только там, где известен объём работы. Пауза останавливает продолжение этапа; текущий запрос к источнику может завершиться.
Akyol Safa, Tok Doğan Sabri, Öğreden Ercan, Demirelli Erhan, Altun Ertürk, Baş Eren, Eğrioğlu Erol, Oğuz Ural
Аннотация
OBJECTIVE: Difficult ureter (DU) is a challenging condition that prevents successful access to the upper urinary tract during retrograde ureteroscopic lithotripsy (URS). The present study aimed to identify clinical risk factors for DU and to evaluate the predictive performance of machine learning-based models in patients undergoing URS for ureteral stones.
METHODS: Medical records of patients who underwent retrograde ureteroscopic lithotripsy between January 2023 and January 2024 were retrospectively reviewed. A total of 221 patients were included and classified into difficult ureter (DU) and non-difficult ureter (NDU) groups. Demographic characteristics, comorbidities, laboratory parameters, and perioperative findings were compared. Conventional statistical analyses were performed using univariate and multivariate regression models. In addition, supervised machine learning algorithms, including decision tree, logistic regression, naive Bayes, support vector machine, k-nearest neighbors, bagged trees, and multilayer perceptron artificial neural network (ANN), were developed to predict DU. Model performance was assessed using 3-fold cross-validation.
RESULTS: Difficult ureter was identified in 34 patients (15.4%). Absence of hydronephrosis, lack of medical expulsive therapy, elevated C-reactive protein (CRP > 5 mg/L), and gout were significantly associated with DU. Multivariate logistic regression analysis identified elevated CRP levels and diabetes mellitus as independent predictors. Among the machine learning models, the multilayer perceptron ANN demonstrated the best performance, with an accuracy of 86.9% and an area under the receiver operating characteristic curve of 0.816. Feature importance analysis revealed CRP as the most influential predictor, consistent with conventional statistical findings.
CONCLUSION: Elevated CRP levels, gout, absence of hydronephrosis, and lack of medical expulsive therapy are important risk factors for difficult ureter during ureteroscopic lithotripsy. Machine learning-based models, particularly artificial neural networks, may provide additional predictive value and serve as decision-support tools for preoperative risk stratification in clinical practice.
METHODS: Medical records of patients who underwent retrograde ureteroscopic lithotripsy between January 2023 and January 2024 were retrospectively reviewed. A total of 221 patients were included and classified into difficult ureter (DU) and non-difficult ureter (NDU) groups. Demographic characteristics, comorbidities, laboratory parameters, and perioperative findings were compared. Conventional statistical analyses were performed using univariate and multivariate regression models. In addition, supervised machine learning algorithms, including decision tree, logistic regression, naive Bayes, support vector machine, k-nearest neighbors, bagged trees, and multilayer perceptron artificial neural network (ANN), were developed to predict DU. Model performance was assessed using 3-fold cross-validation.
RESULTS: Difficult ureter was identified in 34 patients (15.4%). Absence of hydronephrosis, lack of medical expulsive therapy, elevated C-reactive protein (CRP > 5 mg/L), and gout were significantly associated with DU. Multivariate logistic regression analysis identified elevated CRP levels and diabetes mellitus as independent predictors. Among the machine learning models, the multilayer perceptron ANN demonstrated the best performance, with an accuracy of 86.9% and an area under the receiver operating characteristic curve of 0.816. Feature importance analysis revealed CRP as the most influential predictor, consistent with conventional statistical findings.
CONCLUSION: Elevated CRP levels, gout, absence of hydronephrosis, and lack of medical expulsive therapy are important risk factors for difficult ureter during ureteroscopic lithotripsy. Machine learning-based models, particularly artificial neural networks, may provide additional predictive value and serve as decision-support tools for preoperative risk stratification in clinical practice.