biostudies-literatureMa Hthe Fundamental Research Program of Shanxi ProvinceScientific Research Project Plan of Shanxi Provincial Health CommissionAgreement on Horizontal Scientific Research Cooperation Project of the First Hospital of Shanxi Medical University1597https://www.ebi.ac.uk/biostudies/studies/S-EPMC12532899biostudies-literatureUnknown25(1)<h4>Background</h4>Accurate prediction of prognosis and risk stratification in patients with laryngeal cancer can inform appropriate treatment decision-making. This study aims to develop a multi-channel deep learning radiomics model based on contrast-enhanced computed tomography (CECT) for predicting postoperative overall survival (OS) in patients.<h4>Methods</h4>A total of 272 patients with laryngeal cancer were retrospectively recruited from two hospitals between January 2016 and July 2021. Specifically, 156 patients were enrolled from Center 1 as the training cohort, and 116 patients from Center 2 as the external test cohort. Two imaging signatures, reflecting phenotypes of the radiomics and multi-channel deep learning features, were constructed using pretreatment venous-phase CECT imagBMC cancerMulti-channel deep learning radiomics model based on contrast-enhanced CT for predicting postoperative prognosis in laryngeal carcinoma.PMC12532899Nos 2024056Nos. 20210302123253Nos. 20210302123256Nos. 08983Li ZZhang JZhang HMa HLiang LWang WWei WZhang LHao YZhang QWang LfalseMulti-channel deep learning radiomics model based on contrast-enhanced CT for predicting postoperative prognosis in laryngeal carcinoma.<h4>Background</h4>Accurate prediction of prognosis and risk stratification in patients with laryngeal cancer can inform appropriate treatment decision-making. This study aims to develop a multi-channel deep learning radiomics model based on contrast-enhanced computed tomography (CECT) for predicting postoperative overall survival (OS) in patients.<h4>Methods</h4>A total of 272 patients with laryngeal cancer were retrospectively recruited from two hospitals between January 2016 and July 2021. Specifically, 156 patients were enrolled from Center 1 as the training cohort, and 116 patients from Center 2 as the external test cohort. Two imaging signatures, reflecting phenotypes of the radiomics and multi-channel deep learning features, were constructed using pretreatment venous-phase CECT imag2025-01-01T00:00:00Z2025 Oct2026-06-04T13:44:15.456Z2026-05-09T03:12:23.834ZS-EPMC125328994110269410.1186/s12885-025-14912-4