Table 3.
Risk factors for invasiveness of the ground-glass nodule (GGN) in the training set.
Univariate logistic regression | Multivariate logistic regression | |||
---|---|---|---|---|
Characteristics | OR (95%CI) | p | OR (95%CI) | p |
Intra-nodular model | ||||
Rad-score | 5.027 (2.500–10.105) | <0.001 | NA | NA |
Peri-nodular model | ||||
Rad-score | 7.148 (2.722–18.771) | <0.001 | NA | NA |
Gross model | ||||
Rad-score | 5.435 (2.548–11.429) | <0.001 | 14.420 (3.700–56.180) | <0.001 |
Clinical model | ||||
Age | 1.056 (1.011–1.104) | 0.015 | NA | NA |
Diameter | 1.149 (1.037–1.273) | 0.008 | 0.800 (0.605–0.980) | 0.031 |
Spiculation | 4.800 (1.599–14.410) | 0.005 | NA | NA |
Lobulation | 3.200 (1.231–8.317) | 0.016 | NA | NA |
Vessel convergence | 3.542 (1.225–10.236) | 0.019 | NA | NA |
CI, confidence interval; GGN, ground-glass nodule; OR, odds ratio; NA, not available.