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, Thomas R W Herrmann, Olivier Traxer, Bhaskar K Somani
{"title":"人工智能用于结石疾病输尿管镜检查并发症的预测因素和相关性:来自FLEXOR登记的结果","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, Thomas R W Herrmann, Olivier Traxer, Bhaskar K Somani","doi":"10.1007/s00240-025-01763-8","DOIUrl":null,"url":null,"abstract":"<p><p>We aimed to develop machine learning(ML) algorithms to evaluate complications of flexible ureteroscopy and laser lithotripsy(fURSL), providing a valid predictive model. 15 ML algorithms were trained on a large number fURSL data from > 6500 patients from the international FLEXOR database. fURSL complications included pelvicalyceal system(PCS) bleeding, ureteric/PCS injury, fever and sepsis. Pre-treatment characteristics served as input for ML training and testing. Correlation and logistic regression analysis were carried out by a multi-task neural network, while explainable AI was used for the predictive model. ML algorithms performed excellently. For intraoperative PCS bleeding, Extra Tree Classifier achieved the best accuracy at 95.03% (precision 80.99%), and greatest correlation with stone diameter(0.21) and residual fragments(0.26). PCS injury was best predicted by RandomForest (accuracy 97.72%, precision 63.50%). XGBoost performed best for ureteric injury (accuracy 96.88%, precision 60.67%). Both demonstrated moderate correlation with preoperative characteristics. Postoperative fever was predicted by Extra Tree Classifier with 91.34% accuracy (precision 58.20%). Cat Boost Classifier predicted postoperative sepsis with 99.15% accuracy (precision 66.38%), and the best overall performance. At logistic regression, postoperative fever/sepsis positively correlated with preoperative urine culture(p = 0.001). ML represents a powerful tool for automatic prediction of outcomes. Our study showed promises in algorithms training and validation on a very large database of patients treated for urolithiasis, with excellent accuracy for prediction of complications. With further research, reliable predictive nomograms could be created based on ML analysis, to serve as aid to urologists and patients in the decision making and treatment planning process.</p>","PeriodicalId":23411,"journal":{"name":"Urolithiasis","volume":"53 1","pages":"89"},"PeriodicalIF":2.0000,"publicationDate":"2025-05-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12078356/pdf/","citationCount":"0","resultStr":"{\"title\":\"Predictors and associations of complications in ureteroscopy for stone disease using AI: outcomes from the FLEXOR registry.\",\"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, Thomas R W Herrmann, Olivier Traxer, Bhaskar K Somani\",\"doi\":\"10.1007/s00240-025-01763-8\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"<p><p>We aimed to develop machine learning(ML) algorithms to evaluate complications of flexible ureteroscopy and laser lithotripsy(fURSL), providing a valid predictive model. 15 ML algorithms were trained on a large number fURSL data from > 6500 patients from the international FLEXOR database. fURSL complications included pelvicalyceal system(PCS) bleeding, ureteric/PCS injury, fever and sepsis. Pre-treatment characteristics served as input for ML training and testing. Correlation and logistic regression analysis were carried out by a multi-task neural network, while explainable AI was used for the predictive model. ML algorithms performed excellently. For intraoperative PCS bleeding, Extra Tree Classifier achieved the best accuracy at 95.03% (precision 80.99%), and greatest correlation with stone diameter(0.21) and residual fragments(0.26). PCS injury was best predicted by RandomForest (accuracy 97.72%, precision 63.50%). XGBoost performed best for ureteric injury (accuracy 96.88%, precision 60.67%). Both demonstrated moderate correlation with preoperative characteristics. Postoperative fever was predicted by Extra Tree Classifier with 91.34% accuracy (precision 58.20%). Cat Boost Classifier predicted postoperative sepsis with 99.15% accuracy (precision 66.38%), and the best overall performance. At logistic regression, postoperative fever/sepsis positively correlated with preoperative urine culture(p = 0.001). ML represents a powerful tool for automatic prediction of outcomes. Our study showed promises in algorithms training and validation on a very large database of patients treated for urolithiasis, with excellent accuracy for prediction of complications. With further research, reliable predictive nomograms could be created based on ML analysis, to serve as aid to urologists and patients in the decision making and treatment planning process.</p>\",\"PeriodicalId\":23411,\"journal\":{\"name\":\"Urolithiasis\",\"volume\":\"53 1\",\"pages\":\"89\"},\"PeriodicalIF\":2.0000,\"publicationDate\":\"2025-05-14\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12078356/pdf/\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Urolithiasis\",\"FirstCategoryId\":\"3\",\"ListUrlMain\":\"https://doi.org/10.1007/s00240-025-01763-8\",\"RegionNum\":2,\"RegionCategory\":\"医学\",\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"Q2\",\"JCRName\":\"UROLOGY & NEPHROLOGY\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Urolithiasis","FirstCategoryId":"3","ListUrlMain":"https://doi.org/10.1007/s00240-025-01763-8","RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q2","JCRName":"UROLOGY & NEPHROLOGY","Score":null,"Total":0}
引用次数: 0
摘要
我们的目标是开发机器学习(ML)算法来评估柔性输尿管镜和激光碎石(fURSL)的并发症,提供一个有效的预测模型。15 ML算法在来自国际FLEXOR数据库的bbbb6500例患者的大量fURSL数据上进行了训练。fURSL并发症包括盆腔肾盂系统(PCS)出血、输尿管/PCS损伤、发热和败血症。预处理特征作为ML训练和测试的输入。通过多任务神经网络进行相关性和逻辑回归分析,而可解释的人工智能用于预测模型。ML算法表现出色。对于术中PCS出血,Extra Tree Classifier准确率最高,为95.03%(精密度为80.99%),与结石直径(0.21)、残留碎片(0.26)相关性最高。随机森林预测PCS损伤效果最佳(正确率97.72%,精密度63.50%)。XGBoost治疗输尿管损伤效果最佳(正确率96.88%,精密度60.67%)。两者均显示与术前特征有中度相关性。Extra Tree Classifier预测术后发热的准确率为91.34%(精密度为58.20%)。Cat Boost分类器预测术后脓毒症的准确率为99.15%(精密度为66.38%),综合性能最佳。经logistic回归分析,术后发热/败血症与术前尿培养呈正相关(p = 0.001)。ML是自动预测结果的强大工具。我们的研究在一个非常大的尿石症患者数据库上显示了算法训练和验证的前景,在预测并发症方面具有很高的准确性。通过进一步的研究,可以基于ML分析创建可靠的预测图,以帮助泌尿科医生和患者在决策和治疗计划过程中。
Predictors and associations of complications in ureteroscopy for stone disease using AI: outcomes from the FLEXOR registry.
We aimed to develop machine learning(ML) algorithms to evaluate complications of flexible ureteroscopy and laser lithotripsy(fURSL), providing a valid predictive model. 15 ML algorithms were trained on a large number fURSL data from > 6500 patients from the international FLEXOR database. fURSL complications included pelvicalyceal system(PCS) bleeding, ureteric/PCS injury, fever and sepsis. Pre-treatment characteristics served as input for ML training and testing. Correlation and logistic regression analysis were carried out by a multi-task neural network, while explainable AI was used for the predictive model. ML algorithms performed excellently. For intraoperative PCS bleeding, Extra Tree Classifier achieved the best accuracy at 95.03% (precision 80.99%), and greatest correlation with stone diameter(0.21) and residual fragments(0.26). PCS injury was best predicted by RandomForest (accuracy 97.72%, precision 63.50%). XGBoost performed best for ureteric injury (accuracy 96.88%, precision 60.67%). Both demonstrated moderate correlation with preoperative characteristics. Postoperative fever was predicted by Extra Tree Classifier with 91.34% accuracy (precision 58.20%). Cat Boost Classifier predicted postoperative sepsis with 99.15% accuracy (precision 66.38%), and the best overall performance. At logistic regression, postoperative fever/sepsis positively correlated with preoperative urine culture(p = 0.001). ML represents a powerful tool for automatic prediction of outcomes. Our study showed promises in algorithms training and validation on a very large database of patients treated for urolithiasis, with excellent accuracy for prediction of complications. With further research, reliable predictive nomograms could be created based on ML analysis, to serve as aid to urologists and patients in the decision making and treatment planning process.
期刊介绍:
Official Journal of the International Urolithiasis Society
The journal aims to publish original articles in the fields of clinical and experimental investigation only within the sphere of urolithiasis and its related areas of research. The journal covers all aspects of urolithiasis research including the diagnosis, epidemiology, pathogenesis, genetics, clinical biochemistry, open and non-invasive surgical intervention, nephrological investigation, chemistry and prophylaxis of the disorder. The Editor welcomes contributions on topics of interest to urologists, nephrologists, radiologists, clinical biochemists, epidemiologists, nutritionists, basic scientists and nurses working in that field.
Contributions may be submitted as full-length articles or as rapid communications in the form of Letters to the Editor. Articles should be original and should contain important new findings from carefully conducted studies designed to produce statistically significant data. Please note that we no longer publish articles classified as Case Reports. Editorials and review articles may be published by invitation from the Editorial Board. All submissions are peer-reviewed. Through an electronic system for the submission and review of manuscripts, the Editor and Associate Editors aim to make publication accessible as quickly as possible to a large number of readers throughout the world.