Objective To explore the value of radiomics model based on spectral CT to predict lymph node metastasis in clinical stage T1 lung adenocarcinoma.Methods 154 patients with clinical stage of T1 lung adenocarcinoma confirmed by pathological diagnosis were analyzed retrospectively.Based on the images of 70 keV single-energy CT values,the least absolute shrinkage and selection operator(LASSO)and multivariate Logistic regression analysis were used to select the radiomics features for constructing radiomics model;several spectral parameters of primary tumor were measured,then clinical model was established by combining the spectral parameters,clinical features and CT features of patients;the comprehensive model was constructed based on the radiomics model and clinical model and draw the nomogram.Results Receiver operating characteristic(ROC)curve showed that the comprehensive model presented the highest predictive efficiency in the training set and test set,and the area under the curve(AUC)values were 0.94(95％Cl:0.89-0.99)and 0.79(95％Cl:0.65-0.92),respectively.The results of DCA showed that the clinical utility of the nomogram was higher than that in radiomics model and clinical model within a certain threshold range.Conclusion The radiomics model based on spectral CT has a good predictive value for lymph node metastasis in clinical stage T1 lung adenocarcinoma.
维普
万方数据