Growth Prediction of Ground-Glass Nodules Based on Pulmonary Vascular Morphology Nomogram.

Abstract

Rationale and objectives: To construct a nomogram combining conventional CT features (CCTFs), morphologically abnormal tumor-related vessels (MATRVs), and clinical features to predict the two-year growth of lung ground-glass nodule (GGN).

Methods: High-resolution CT targeted scan images of 158 patients including 167 GGNs from January 2016 to September 2019 were retrospectively analyzed. The CCTF and MATRV of each GGN were recorded. All GGNs were randomly divided into a training set (n = 118) and a validation set (n = 49). Multiple stepwise regression was used to select the features. Multivariate logistic regression was used to construct the CCTF, CCTF-CTRV (category of tumor-related vessel), and CCTF-QTRV (quantity of tumor-related vessel) nomograms. The performance and utility of the nomograms were evaluated using the area under the receiver operating characteristic curve (AUC), net reclassification improvement (NRI), integrated discrimination improvement (IDI), and decision curve analysis (DCA).

Results: The AUC of the CCTF-QTRV nomogram, which included the features of smoking history, nodule pattern, lobulation, and the number of distorted and dilated vessels, was higher than the AUCs of the CCTF and CCTF-CTRV nomograms in both the training set (AUC: 0.906 vs. 0.857; vs. 0.851) and the validation set (AUC: 0.909 vs. 0.796; vs. 0.871). DCA indicated that both patients and clinicians could benefit from using the nomogram.

Conclusion: The nomogram constructed by combining MATRV, CCTF, and clinical information can more effectively predict the two-year growth of GGNs. This integrated approach enhances the predictive accuracy, making it a valuable tool for clinicians in managing and monitoring patients with GGNs.