A Feasibility Study of Classification between Lung Adenocarcinoma Nodules and Inflammatory Nodules Using Intranodal and Perinodular Radiomics Features Based on Spectral CT
Objective To investigate the diagnostic value of dual detector spectral CT parameters and intranodal and perinodal radiomics features in lung adenocarcinoma nodules and inflammatory nodules.Methods The clinical data and spectral CT parameters of 83 patients with lung adenocarcinoma nodules and 62 patients with inflammatory nodules were collected retrospectively.The cases were randomly divided into training set(n=95)and verification set(n=50).The radiomics features were extracted from the 40keV monoenergy images in arterial phase and venous phase by using 3D-Slicer.The focus was manually sketched and named as intranodal ROI,and then expanded 5mm outward to form perinodal ROI using semi-automatic segmentation program.After calculating the intra-group correlation coefficient(ICC),Spelman correlation coefficient,minimum absolute contraction and selection operator algorithm(LASSO)and Logistic analysis were used for feature screening.Logistic regression analysis was used to establish a clinical model,a radiomics model and a combined model.The area under curve(AUC)was calculated to evaluate the performance of the model,and the decision curve analysis(DCA)was used to evaluate the clinical practicability of the model.Results Male,burr sign,NIC and λHU in venous phase were independent clinical predictors of lung adenocarcinoma.And 7 radiomics features were used to construct radiomics models to distinguish lung adenocarcinoma nodules from inflammatory nodules(3 intranodal radiomics features,1 perinodal radiomics features and 3 intranodal and perinodal radiomics features).The combined model is best,AUC is 0.91.Conclusion Spectral CT parameters,intranodal and perinodular radiomics features can be used to distinguish lung adenocarcinoma nodules from inflammatory nodules.
万方数据
维普
