Skip to main content
. Author manuscript; available in PMC: 2020 Aug 15.
Published in final edited form as: Expert Syst Appl. 2019 Jan 18;128:84–95. doi: 10.1016/j.eswa.2019.01.048

Table 6:

Comparison with other current deep learning models.

Method Nodules Hold-out Test Sensitivity Specificity Accuracy AUC
Deep Auto-encoder
(Kumar et al., 2015)		 4,323 No - - 0.7501 -
CNN
(Hua et al., 2015)		 2,545 No 0.737 0.787 - -
DBN
(Hua et al., 2015)		 2,545 No 0.734 0.822 - -
Multi-scale CNN
(Shen et al., 2015b)		 1,375 No - - 0.84 -
Multi-crop CNN
(Shen et al., 2017b)		 1,375 No - - 0.8714 -
Proposed 4,252 Yes 0.705 0.889 0.842 0.856