Hi James, If you draw in 255 voxels on the wm.mgz to connect that stray blob of red voxels to the others, then the wm surf should include it on a rerun. Look through a couple slices in that area for any disconnected wm voxels since it may be occuring on multiple slices. -Louis

On Thu, 3 Jul 2014, james pardon wrote:

Hi,  I'm working on some subjects with gray matter lesions, where usually these lesions happen at the boundary of GM and WM (mostly patients with MS). The problem is that Freesurfer can not follow the cortex past this areas.

I have tried to use control points, white matter edits and even editing the problematic regions on orig.mgz volume with neighboring voxels and re running recon-all. My results have improved substantially, however I still have a problem where after white matter edit, and adding some new voxels to wm.mgz, the wm surface (and overlying pial surface) does not respect the wm.mgz. In the attached screenshot for example, I can see some areas that are part of wm.mgz (in red) but not included in WM surface. I was wondering whether anyone can give a hint on what recon-all steps I might need to re-run?

Thanks  James