. Author manuscript; available in PMC: 2020 Mar 1.

Published in final edited form as: IEEE Trans Cybern. 2018 Feb 8;49(3):1123–1136. doi: 10.1109/TCYB.2018.2797905

TABLE I.

Segmentation Performance in Terms of Dice Ratio and Standard Deviation, Achieved by the Baseline Comparison Methods and Our CC-3-D-FCN on 11 Subjects. The Highest Performance in Each Tissue Class Is Highlighted in Bold

	WM	GM	CSF
FAST	.4641(.0791)	.4045(.0979)	.5636(.2334)
MV	.5729(.0327)	.7099(.0412)	.7160(.0386)
RF	.8236(.0164)	.8420(.0126)	.8607(.0170)
LINKS	.8424(.0183)	.8652(.0154)	.8896(.0251)
DeepMedic [54]	.8458(.0150)	.8608(.0158)	.9196(.0210)
3D-UNet [55]	.8418(.0161)	.8680(.0172)	.9012(.0225)
CC-2D-FCN	.7583(.0215)	.8146(.0115)	.8586(.0101)
3D-CNN	.8262(.0242)	.8413(0.0329)	.8999(.0269)
CC-3D-FCN	.8586(.0139)	.8817(.0158)	.9269(.0201)