{"title":"The Predictive Value of Radiomics for Esophagotracheal Fistula after Radiotherapy in Esophageal Cancer.","authors":"Huiyao Chen, Yanglong Wu, Congcong Wu","doi":"10.2174/0115734056452685260414044447","DOIUrl":null,"url":null,"abstract":"<p><strong>Introduction: </strong>Esophagotracheal Fistula (ETF) is a serious complication following radiotherapy for esophageal cancer, with treatment outcomes significantly worse than expected.</p><p><strong>Methods: </strong>Pre-radiotherapy CT images and clinical data from patients with esophageal malignancies treated at the Second Affiliated Hospital of Wenzhou Medical University between January 2015 and December 2023 were retrospectively analyzed. Tumor contours were manually delineated using 3D Slicer, and radiomic features were extracted using PyRadiomics. Features associated with ETF development (p<0.05) were identified via the Mann-Whitney U test and further refined using Least Absolute Shrinkage and Selection Operator (LASSO) regression to determine the final radiomic signature. Subsequently, univariate and multivariate binary logistic regression analyses were performed.</p><p><strong>Results: </strong>The study included 77 patients, 30 of whom developed ETF. Of the initial 845 radiomic features, 10 were significantly associated with ETF. Among clinical factors, the type of radiation therapy was an independent predictor for ETF. In the training cohort, the radiomics model achieved an AUC of 0.866 (95% CI: 0.7907-0.9402), with a sensitivity of 0.831 and specificity of 0.792. The combined model (radiomics + clinical features) achieved an AUC of 0.892 (95% CI: 0.8238-0.9601), sensitivity of 0.823, and specificity of 0.912. In the validation cohort, the radiomics model had an AUC of 0.736 (95% CI: 0.5781-0.8947), sensitivity of 0.833, and specificity of 0.621. The combined model achieved an AUC of 0.791 (95% CI: 0.6461-0.9354), sensitivity of 0.822, and specificity of 0.797.</p><p><strong>Discussion: </strong>The combination of radiomic and clinical features achieves excellent AUC performance and shows potential for the non-invasive prediction of ETF following radiotherapy in esophageal cancer patients.</p><p><strong>Conclusion: </strong>The model, combined with radiomic and clinical features, has great predictive value.</p>","PeriodicalId":54215,"journal":{"name":"Current Medical Imaging Reviews","volume":" ","pages":""},"PeriodicalIF":1.1000,"publicationDate":"2026-04-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Current Medical Imaging Reviews","FirstCategoryId":"3","ListUrlMain":"https://doi.org/10.2174/0115734056452685260414044447","RegionNum":4,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q3","JCRName":"RADIOLOGY, NUCLEAR MEDICINE & MEDICAL IMAGING","Score":null,"Total":0}
引用次数: 0
Abstract
Introduction: Esophagotracheal Fistula (ETF) is a serious complication following radiotherapy for esophageal cancer, with treatment outcomes significantly worse than expected.
Methods: Pre-radiotherapy CT images and clinical data from patients with esophageal malignancies treated at the Second Affiliated Hospital of Wenzhou Medical University between January 2015 and December 2023 were retrospectively analyzed. Tumor contours were manually delineated using 3D Slicer, and radiomic features were extracted using PyRadiomics. Features associated with ETF development (p<0.05) were identified via the Mann-Whitney U test and further refined using Least Absolute Shrinkage and Selection Operator (LASSO) regression to determine the final radiomic signature. Subsequently, univariate and multivariate binary logistic regression analyses were performed.
Results: The study included 77 patients, 30 of whom developed ETF. Of the initial 845 radiomic features, 10 were significantly associated with ETF. Among clinical factors, the type of radiation therapy was an independent predictor for ETF. In the training cohort, the radiomics model achieved an AUC of 0.866 (95% CI: 0.7907-0.9402), with a sensitivity of 0.831 and specificity of 0.792. The combined model (radiomics + clinical features) achieved an AUC of 0.892 (95% CI: 0.8238-0.9601), sensitivity of 0.823, and specificity of 0.912. In the validation cohort, the radiomics model had an AUC of 0.736 (95% CI: 0.5781-0.8947), sensitivity of 0.833, and specificity of 0.621. The combined model achieved an AUC of 0.791 (95% CI: 0.6461-0.9354), sensitivity of 0.822, and specificity of 0.797.
Discussion: The combination of radiomic and clinical features achieves excellent AUC performance and shows potential for the non-invasive prediction of ETF following radiotherapy in esophageal cancer patients.
Conclusion: The model, combined with radiomic and clinical features, has great predictive value.
期刊介绍：
Current Medical Imaging Reviews publishes frontier review articles, original research articles, drug clinical trial studies and guest edited thematic issues on all the latest advances on medical imaging dedicated to clinical research. All relevant areas are covered by the journal, including advances in the diagnosis, instrumentation and therapeutic applications related to all modern medical imaging techniques.
The journal is essential reading for all clinicians and researchers involved in medical imaging and diagnosis.
求助内容：
应助结果提醒方式：
京公网安备 11010802042870号
