Scholarlink

<p>The aim of this study was to identify an optimal model for predicting the short-term efficacy of ultrasound-guided extracorporeal shock wave lithotripsy (SWL) utilizing clinical and radiomic parameters. Patients requiring SWL for ureteral stones between June 1, 2021, and November 30, 2024, were consecutively divided into training, validation and testing cohorts. Radiomic features of regions depicting stones were extracted from computed tomography images using PyRadiomics. To identify independent predictors for the short-term success of SWL, feature selection and univariate and multivariate analyses were conducted. Additionally, models were created by combining radiomic features with clinical predictors and evaluated through receiver operating characteristic (ROC) curve analysis. Calibration and decision curve analyses were employed to assess the fit and clinical benefit of the models. In total, 218 patients were enrolled in the training, validation, and testing cohorts, comprising 125, 42, and 51 patients, respectively. Following logistic regression analyses and model development, ROC curve analysis indicated that the area under the curve values for the clinical, radiomic, and combined models were 0.757, 0.860, and 0.861 for the training group; 0.739, 0.861, and 0.852 for the validation group; and 0.731, 0.817 and 0.817 for the test group, respectively. The calibration curves demonstrated a good fit of the combined model in both cohorts. Decision curve analysis indicated greater net clinical benefits for both the combined and radiomic models than for the clinical model. In conclusion, compared with the combined model, the radiomic model demonstrated comparable predictive performance, and both outperformed the clinical model in predicting short-term SWL efficacy. These findings may have practical implications for treatment decisions.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Clinicoradiomic analysis for predicting short-term treatment efficacy after one session of ultrasound-guided extracorporeal wave lithotripsy in patients with a single ureteral stone

  • 期刊论文

  • 发表日期：

    2026年4月21日

查看全文
  • Yueyue Zhang,
  • Meng Wang,
  • Chenyi He,
  • Hua Tong,
  • Bin Liu,
  • Jueqi Chen

摘要

The aim of this study was to identify an optimal model for predicting the short-term efficacy of ultrasound-guided extracorporeal shock wave lithotripsy (SWL) utilizing clinical and radiomic parameters. Patients requiring SWL for ureteral stones between June 1, 2021, and November 30, 2024, were consecutively divided into training, validation and testing cohorts. Radiomic features of regions depicting stones were extracted from computed tomography images using PyRadiomics. To identify independent predictors for the short-term success of SWL, feature selection and univariate and multivariate analyses were conducted. Additionally, models were created by combining radiomic features with clinical predictors and evaluated through receiver operating characteristic (ROC) curve analysis. Calibration and decision curve analyses were employed to assess the fit and clinical benefit of the models. In total, 218 patients were enrolled in the training, validation, and testing cohorts, comprising 125, 42, and 51 patients, respectively. Following logistic regression analyses and model development, ROC curve analysis indicated that the area under the curve values for the clinical, radiomic, and combined models were 0.757, 0.860, and 0.861 for the training group; 0.739, 0.861, and 0.852 for the validation group; and 0.731, 0.817 and 0.817 for the test group, respectively. The calibration curves demonstrated a good fit of the combined model in both cohorts. Decision curve analysis indicated greater net clinical benefits for both the combined and radiomic models than for the clinical model. In conclusion, compared with the combined model, the radiomic model demonstrated comparable predictive performance, and both outperformed the clinical model in predicting short-term SWL efficacy. These findings may have practical implications for treatment decisions.