Skip to main content
. Author manuscript; available in PMC: 2014 Aug 19.
Published in final edited form as: Neuroimage. 2014 Mar 18;94:65–78. doi: 10.1016/j.neuroimage.2014.03.026

Fig. 3.

Fig. 3

Top: Mean FA image of 60 subjects coregistered by the FSL TBSS pipeline and masked in its standard method of removing all voxels that are zero in at least one subject. Bottom: The same image, masked in the proposed method of removing only voxels that are zero in more than half of subjects. Note that “holes” in WM are prevented, and much more of the midbrain is left intact.