A novel localization technique for peripheral ground glass opacity using geometric parameters measured on CT images

Table 4 A univariate linear regression analysis of the GGO related characteristics associated with the distance from nodule to the marked visceral pleura and a univariate logistic regression analysis of the GGO related characteristics associated with the sufficient resection margin

	Univariate linear regression analysis (The distance from nodule to the marked visceral pleura)			Univariate logistic regression analysis (sufficient resection margin)
	Coefficient	t	P	Odds ratio	95% CI	P
Lesion size	0.008	0.971	0.334	0.969	0.678,1.385	0.864
Lesion type	− 0.037	− 0.503	0.616	–	–	0.998
Lesion location	0.007	0.303	0.762	0.932	0.356,2.438	0.886
X-axis value	0.061	21.233	0.000	0.875	0.638,1.199	0.405
Y-axis value	− 0.002	− 0.108	0.914	1.216	0.714,2.071	0.471
Z-axis value	− 0.027	− 0.450	0.653	12.204	1.749,85.137	0.012

X-axis value: the measured distance between the parietal pleural projection of GGO and landmark (capitulum costae or sternocostal joint) on CT images; Y-axis value: Y-axis value referred to the rib or intercostal space corresponding to the parietal pleural projection; Z-axis value: lesion depth from pleura
P < 0.20 was considered as significant which would be included in the multivariate regression analysis
CI confidence interval

ISSN: 1471-2482