Dear Mel,

if your recon in 5.3 version finished without error, there must be files lh.orig.nofix and rh.orig.nofix present in the subj_id/surf directory. This type of error is most frequently due to incorrectly fixed topological defect. The original wm.mgz and orig.nofix errors can be quite a large distance from the area of your white surface error. The fix you are referring to (editing the wm.mgz) makes perfect sense in this context.

Antonin

That does not make sense. Can you run this subject in v6? If the problem goes away, then there is not much we can do about it in 5.3 On 02/28/2017 05:27 AM, Melanie Ganz wrote:

Hi Doug,

the orig looks fine and we cannot see a file orig.nofix in the subject dirs. We still use v5.3.

But what we tried now for another similar subject was to remove falsely labelled white matter in dura which was far away, but that fixed the white matter errors so far inside the brain. Does that make any sense to you why that helps?

Cheers, Mel

On 27-02-2017 5:23, Douglas Greve wrote:

...

Hi Mel, how does the orig or orig.nofix look?

On 2/24/17 9:21 AM, Melanie Ganz wrote:

...

Dear list,

we are experiencing a strange wm surface error that we are not quite sure about how to fix. The errors is at a location where the wm.mgz look absolutely fine as well as the brainmask.mgz. Also the intensities in the norm.mgz look fine.

Please find 6 pics showing the error highlighted by the cursor on the brainmask.mgz and the wm.mgz in all three views.

Cheers, Mel