基于18F-FDG PET/CT的CT影像组学模型检测腹膜恶性病变的价值OA

Objective To investigate the value of a CT radiomics model based on 18F-FDG PET/CT in detecting peritoneal malignant lesions.Methods Clinical data of 98 patients with suspected peritoneal malignant lesions who underwent whole-body 18F-FDG PET/CT examination and had pathological results were retrospectively collected.According to pathological results,the patients were divided into a peritoneal malignant lesion group(58 cases)and a non-malignant lesion group(40 cases).Pearson correlation coefficients and least absolute shrinkage and selection operator(LASSO)regression were used to select CT radiomics features,and a support vector machine(SVM)algorithm was used to establish the radiomics model.Five-fold cross-validation was performed for model training and validation.Clinical data,CT features,PET/CT parameters,and radiomics score(Radscore)with statistically significant differences between the two groups were analyzed using multivariate logistic regression to identify independent risk factors,and the clinical model,CT model,and PET model were constructed.Receiver operating characteristic(ROC)curve analysis was used to evaluate the diagnostic performance of the models,and the area under the curve(AUC)was calculated.The optimal cutoff value of the radiomics model was obtained,and the sensitivity and specificity of the radiomics model were calculated.The DeLong test was used to compare differences in AUC values among models,and decision curve analysis(DCA)was used to evaluate the clinical net benefit of the models.Results Multivariate logistic regression analysis showed that sex,abdominal distension,anorexia,ascites,peritoneal thickening,SUVmax,SUVmean,and Radscore were independent risk factors for diagnosing peritoneal malignant lesions.The clinical model was constructed using patient sex,abdominal distension,and anorexia;the CT model was constructed using ascites and peritoneal thickening;and the PET model was constructed using SUVmax and SUVmean.The diagnostic performance of the radiomics model was higher than that of the clinical model,CT model,and PET model(all P<0.05).The optimal cutoff value of the radiomics model was 0.296,with sensitivity and specificity of 0.983 and 1.000,respectively.DCA showed that the radiomics model achieved a higher clinical net benefit when the threshold probability was>0.130.Conclusion The CT radiomics model based on 18F-FDG PET/CT demonstrates excellent performance in diagnosing peritoneal malignant lesions and is significantly superior to the clinical model,the CT model based on 18F-FDG PET/CT,and the PET model.