Objective To analyze the role of radiomics in the diagnosis of different pathological subtypes of chronic rhinosinusitis. Methods The clinical data of patients with chronic rhinosinusitis who underwent endoscopic sinus surgery at the First Affiliated Hospital of Kunming Medical University from February 2023 to September 2024 were retrospectively analyzed. The data were screened according to the inclusion and exclusion rules, and finally 143 cases were included. CT imaging data of the disease patients were collected. The cases were randomly divided into a training set (n = 100) and a validation set (n = 43). After extracting deep features of CT images through radiomics technology, the LASSO (Least Absolute Shrinkage and Selection Operator) algorithm was used to screen the features most likely to predict eosinophilic chronic rhinosinusitis. Combined with clinical data, single-factor and multi-factor logistic regression (LR, logistic Regression) algorithms were used to establish radiomics models, clinical models, and combined models. Finally, the clinical application value and predictive efficacy of the models were evaluated by calibration curve analysis (CCA) and decision curve analysis (DCA). Results The area under the curve (AUC) of the radiomics model in the training set was 0.96, and in the test set was 0.80. Compared with the clinical data model, the AUC in the training set was 0.93 and in the test set was 0.77. The predictive effect was better (DeLong test, P < 0.001), and the combined model with clinical data had better efficacy than the individual clinical data model and radiomics model. The AUC in the training set and test set climbed to 0.98 and 0.86, respectively. The predictive performance of these models was proved to have clinical application value by CCA and DCA evaluation. Conclusion The CT radiomics-based model, particularly the combined model integrating clinical data, enables precise subtyping of eosinophilic chronic rhinosinusitis (eCRS). This provides a novel noninvasive assessment for preoperative CRS patients, thus enabling prognosis evaluation and facilitating the development of personalized precision treatment strategies.