This study aimed to develop a model based on radiomics features extracted from computed tomography (CT) images to effectively differentiate between minimally invasive adenocarcinomas (MIAs) and invasive adenocarcinomas (IAs) manifesting as pure ground-glass nodules (pGGNs) larger than 10 mm. Method: This retrospective study included patients who underwent surgical resection for persistent pGGN between November 2012 and June 2018 and diagnosed with MIAs or IAs. The patients were randomly assigned to the training and test cohorts. The correlation coefficient method and the least absolute shrinkage and selection operator (LASSO) method were applied to select radiomics features useful for constructing a model whose performance was assessed by the area under the receiver operating characteristic curve (AUC-ROC). The radiomics model was compared to a standard CT model (shape, volume and mean CT value of the largest crosssection) and the combined radiomics-standard CT model using univariate and multivariate logistic regression analysis.Results: The radiomics model showed better discriminative ability (training AUC, 0.879; test AUC, 0.877) than the standard CT model (training AUC, 0.820; test AUC, 0.828). The combined model (training AUC, 0.879; test AUC, 0.870) did not demonstrate improved performance compared with the radiomics model. Radiomics_score was an independent predictor of invasiveness following multivariate logistic analysis. Conclusions: For pGGNs larger than 10 mm, the radiomics model demonstrated superior diagnostic performance in differentiating between IAs and MIAs, which may be useful to clinicians for diagnosis and treatment selection.
Purpose
To explore the value of radiomics in the identification of lung adenocarcinomas with predominant lepidic growth in pure ground-glass nodules (pGGNs) larger than 10 mm.
Methods
We retrospectively analyzed CT images of 204 patients with large pGGNs (≥ 10 mm) pathologically diagnosed as minimally invasive adenocarcinomas (MIAs), lepidic predominant adenocarcinomas (LPAs), and non-lepidic predominant adenocarcinomas (NLPAs). All pGGNs in the two groups (MIA/LPA and NLPA) were randomly divided into training and test cohorts. Forty-seven patients from another center formed the external validation cohort. Baseline features, including clinical data and CT morphological and quantitative parameters, were collected to establish a baseline model. The radiomics model was built with the optimal radiomics features. The combined model was developed using the rad_score and independent baseline predictors. The performance of the models was evaluated using the area under the receiver operating characteristic curve (AUC) and compared using the DeLong test. The differential diagnosis performance of the models was compared with three radiologists (with 20+, 10+, and 3 years of experience) in the test cohort.
Results
The radiomics (training AUC: 0.833; test AUC: 0.804; and external validation AUC: 0.792) and combined (AUC: 0.849, 0.820, and 0.775, respectively) models performed better for discriminating than the baseline model (AUC: 0.756, 0.762, and 0.725, respectively) developed by tumor location and mean CT value of the whole nodule. The DeLong test showed that the AUCs of the combined and radiomics models were significantly increased in the training cohort. The highest AUC value of the radiologists was 0.600.
Conclusion
The application of CT radiomics improved the identification performance of lung adenocarcinomas with predominant lepidic growth appearing as pGGNs larger than 10 mm.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.