Background  Most of the malignant pulmonary ground-glass nodules (GGN) screened by low-dose CT are lung cancer of stage I, preoperative identification of invasive adenocarcinoma determines lobectomy or sublobectomy.

  Objective  To explore the best diagnostic parameters for assessment of the invasiveness of GGN smaller than 2 cm by comprehensively measuring its CT three-dimensional (3D) parameters, so as to provide evidences for clinic diagnosis.

  Methods  Clinical data about 165 patients who were diagnosed with GGN smaller than 2cm and treated by thoracic surgery in Thoracic Surgery Department of Chinese PLA General Hospital from July 1, 2021 to December 31, 2021 were retrospectively collected, including 45 males and 120 females, with mean age of 53.47 ± 10.86 years. According to the invasiveness, AAH (atypical adenomatous hyperplasia), AIS (adenocarcinoma in situ) and MIA (minimally invasive adenocarcinoma) were included in the low-risk group (n=62), while IA (invasive adenocarcinoma) was included in the high-risk group (n=103). 3D parameters of GGN and its solid components were measured respectively, including diameter, volume, average CT value, density, mass, as well as the ratio of diameter, volume and mass of the solid components to the GGN. The differences of 3D physical parameters between the two groups were analyzed and the risk factors for predicting invasiveness were screened out and their diagnostic efficiency was evaluated.

  Results  Totally 45 males and 120 females were included, with an average age of (53.47 ± 10.86) years. There were 62 cases in the low risk group and 103 cases in the high risk group. There was no significant difference in gender and age between the two groups. The differences of all 3D parameters between the low and high risk groups were all statistically significant (P<0.05). Multiple Logistic regression analysis indicated that patients with larger solid volume (OR=1.017, 95% CI: 1.005-1.029, P=0.007) and longer diameter (OR=1.147, 95% CI: 1.062-1.238, P<0.01) had higher risk of invasive adenocarcinoma. ROC (receiver operating characteristic curve) analysis indicated that the area under curve (AUC, 95% CI) of solid volume and solid diameter were 0.765 (0.566-0.954) and 0.731 (0.500-0.946), their AUC of combined diagnosis was 0.849 (0.701-0.973).

  Conclusion  Diversified CT 3D parameters can assist in the prediction of invasiveness of GGN, especially the combination of solid volume and solid diameter shows a better diagnostic efficiency. It is a new method in addition to the traditional method of measuring the diameter and the average CT value alone.