Table 5.

Performance of various lung nodule segmentation methods on the LIDC-IDRI dataset.

Methods	Year	Nodule amount		Overlap
		Training	Testing
Tachibana and Kido (2006)	2006	–	23	50.7 ± 21.9%
Wang et al. (2009)	2009	23	64	58%
Messay et al. (2010)	2010	–	68	63 ± 16%
Kubota et al. (2011)	2011	–	23	69 ± 18%
			82	59 ± 19%
Tan et al. (2013)	2013	–	23	65%
Lassen et al. (2015)	2015	–	19	52 ± 7%
			40	50 ± 14%
Messay et al. (2015)	2015	300	66	71.70 ± 19.89%
			77	69.23 ± 13.82%
Proposed CF-CNN	2017	350	493	71.16± 12.22%