Abstract
1036
Objectives: Based on 18F-deoxyglucose (FDG) PET/CT to explore the construction and verification of the predictive model for pathologic invasion of early lung adenocarcinoma with ground glass nodules.
Methods: A retrospective analysis was conducted on the PET/CT data on ground-glass nodules (GGNs) patients resected from 149 patients with pre-invasive/minimally invasive adenocarcinoma (MIA) and invasive adenocarcinoma (IAC). The study population is randomly divided into a modeling group and validation group according to 1:1. The differences of qualitative morphological characteristics, quantitative parameters and quantitative functional parameters of pre-invasive lesions/MIA and IAC were compared, respectively. The logistic regression method was used to construct the model, and the receiver operating characteristic (ROC) curve was used to verify the model's robustness.
Results: In the model group, the proportion of mixed GGN, irregular shape, lobulation sign, bronchiectasis/twist/truncation sign, DGGN, DSolid, CTR, CTGGO, SUVmax and SUV index in the IAC group were significantly higher than those in pre-invasion/MIA group (all P < 0.05). Based on the qualitative parameters (GGN type, edge features), quantitative parameters (CTGGO, SUVindex), combined qualitative and quantitative parameters (GGN type, edge diagnosis, SUVindex) of PET/CT, models 1-3 were constructed respectively, and the area under the curve (AUC) of ROC in the modeling group was 0.896, 0.880 and 0.931, respectively. After internal validation, the AUC of model 1 and model 3 decreased significantly in the validation group (0.730 and 0.768, respectively), while the AUC of model 2 did not reduce significantly in the validation group (0.802). Conclusions: The combination of morphological and functional quantitative parameters of 18F-FDG PET/CT can effectively predict the pathological invasion of early lung adenocarcinoma.