[Freesurfer] problem with intensity normalization and mri banding artifacts

Hi all,

I'm running into a problem where some of my scans have varied brightness between slices. What happens is that this results in heavy banding of the resulting scans, and this creates major problem for freesurfer to segment out wm because of the variation in signal intensity.

Intensity normalization doesn't seem to correct this kind of signal problems.

I'm wondering what would be a good way to go about correcting this? I'm been doing control points and wm edits, but because the wm intensity is not standarized across the slices, I think it's just an exercise in futility so far.

I've attached a screen shot.

Thanks!

- David Qixiang Chen