. Author manuscript; available in PMC: 2017 Jul 1.

Published in final edited form as: J Comput Assist Tomogr. 2016 Jul-Aug;40(4):589–595. doi: 10.1097/RCT.0000000000000394

Table 4.

Multiple logistic regression models for distinguishing primary lung cancer from metastases.

Model	Variable	OR^a	95% CI	P value	R²	AUC^b
Clinical	COPD	7.09	1.43–35.2	0.017	0.31	0.77
Clinical	Ever smoked	4.03	1.13–14.4	0.031	0.31	0.77
2D CT	Circularity (unit=0.01)	0.94	0.89–0.98	0.006	0.34	0.81
2D CT	Difference image entropy ^c	0.49	0.02–1.00	0.047	0.34	0.81
3DCT	Lacunarity at box size=32 (unit=0.01)	4.1	1.8–9.3	0.001	0.40	0.83
Clinical-2D CT combined	COPD	10.1	2.04–50.2	0.001	0.43	0.85
Clinical-2D CT combined	Circularity (unit=0.01)	0.92	0.88–0.97	0.005	0.43	0.85
Clinical-3D CT combined	Lacunarity at box size=32 (unit=0.01)	1.10	1.04–1.16	0.001	0.56	0.90
Clinical-3D CT combined	COPD	15.5	2.65–90.5	0.002	0.56	0.90

Odds Ratio (OR) >1 indicates higher likelihood of primary lung cancer as CT parameters increase one unit. Change unit for circularity and lacunarity is 0.01.

Area under receiver operating characteristic curve

Log transformed values