Objective To evaluate the value of CT radiomics in predicting the invasion of pulmonary adenocarcinoma with ground glass nodules.Methods The data of patients with atypical adenomatous hyperplasia,adenocarcinoma in situ,minimally invasive adenocarcinoma and invasive adenocarcinoma confirmed by pathology in the Affiliated Hospital of Nan-tong University from January 2021 to June 2022 were collected.The CT imaging findings were ground glass nodules with a maximum diameter of≤3cm.Among them,atypical adenomatous hyperplasia and adenocarcinoma in situ were non-invasive groups,and minimally invasive adenocarcinoma and invasive adenocarcinoma were invasive groups.A total of 295 patients were included in the study,86 in the non-invasive group and 209 in the invasive group.The training set and validation set were divided according to the ratio of 7∶ 3.The differences of clinical and imaging data between the non-invasive groups and invasive groups were evaluated by chi square test,independent sample t-test and nonparametric rank sum test.3D slicer soft-ware is used to sketch VOI,extract image omics features,and R language software is used to screen features and establish models.Draw the ROC curve,evaluate the predictive performance of the model with the area under the curve(AUC),sen-sitivity and specificity,and evaluate the clinical value of the model with the clinical decision curve analysis(DCA).Re-sults Age,air bronchogram,maximum section length diameter,solid component length and tumor consolidation rate(CTR)were statistically different between the non-invasive group and the invasive group(P<0.05).Multivariate Logistic regres-sion analysis showed that maximum section length diameter was an independent predictor of invasive lung adenocarcinoma(P<0.05).The optimal cut-off value was 11.545mm.A total of 851 features were extracted from CT radiomics,and finally 11 features were included to construct the model.The AUC of CT radiomics model in training set and verification set were 0.902 and 0.93 respectively.Conclusion The model based on CT radiomics has a good efficacy in predicting the inva-sion of lung adenocarcinoma,and is helpful for clinical decision on follow-up management,surgical timing and surgical methods.
NETL
NSTL
万方数据
