Dear Eli,

I have looked at your data.

Due to the fact that the white matter does have intensity well below 110 in the large portion of white matter area under your problematic spot, I think that instead of editing wm.mgz the better way in your case would be to add some control points to the white matter surrounding region

112,49,128.

Btw, it seems that the white matter has generally very inhomogeneus signal intensity. Does not the image originate from the subjects with some kind of severe pathology of white matter?
What is the large hypointense spot centered at 102,64,126? This causes large hole in wm.mgz But there are other hypointensities, for example at 96,77,151.
Does not your problematic spot correspond to some pathology?

Regards,

Antonin


Bruce,

Main problem area that I can see is at coordinates [104,42,124].

My first attempt, I don't think I had the recon editing box checked when
I was editing the white matter, but I used a brush value of 255. My
second attempt I painted over the same voxels but with the recon editing
box checked.

Both times, after saving wm.mgz I ran the following command:
"recon-all -autorecon2-wm -autorecon3 -subjid my_subject -sd `pwd`"
Only thing different from the guide is I specify a different subject
directory.

Below are download links for the important files, let me know if you
would like to look at anything else. If you look at wm.mgz along with
the rh.white, you can see the surface doesn't appear to acknowledge my
edit.

001.mgz:
https://1drv.ms/u/s!AoOsIIayCH7qlxLBNvCT4JCvM1Kc

recon-all.logs.tar.gz:
https://1drv.ms/u/s!AoOsIIayCH7qlxRGOWpUCttk5Zji

brainmask+surf+wm.tar.gz
https://1drv.ms/u/s!AoOsIIayCH7qlxbDSS54gfHkq2M3

Thanks tremendously for your time,
Eli

> ------------------------------
> 
> Message: 4
> Date: Wed, 1 Mar 2017 16:39:33 -0500 (EST)
> From: Bruce Fischl <fis...@nmr.mgh.harvard.edu>
> Subject: Re: [Freesurfer] wm.mgz manual editing
> To: Freesurfer support list <freesurfer@nmr.mgh.harvard.edu>
> Message-ID:
>       
> <https://urldefense.proofpoint.com/v2/url?u=http-3A__alpine.LRH.2.20.1703011638420.14038-40door.nmr.mgh.harvard.edu&d=DwICAg&c=QmPtDiFixEjkMvDKaP3E2Vb9C2z4M0PdarxyAHQ2iDQ&r=xbyrBxm81l6mG_XEX66jgCNTXfK3eVo30T9sVQVTz1Q&m=yBgXIybzSgTt6ZfKp6pNzp_N8aWytMgQmmsZVDx1AwI&s=BM5wWGqXMNLiptKsXqsKDinkU3EtRMoCYFuj0Gbx650&e=
>  >
> Content-Type: text/plain; charset="utf-8"
> 
> Hi Eli
> 
> hmm, maybe you should just upload the subject and email us the voxel 
> coords you are talking about?
> 
> Also, in general it is hard for us to help you unless you include more 
> information. What exact command did you run? What was the screen output? 
> What was in the recon-all.log?
> 
> cheers
> Bruce
> 
> 
> On Wed, 1 
> Mar 2017, Rockers, Elijah D. wrote:
> 
> > Couple of questions:
> >
> > 1)
> >
> > I edited wm.mgz following the instructions in the wiki guide, but it did
> > not change anything after reconstruction. I am still using 5.3 right
> > now, so maybe the guide is wrong for me. What is the brush value
> > supposed to be? I used 255 but it did not extend the white matter
> > surface to include my edits. I checked to make sure they saved
> > correctly, they did, but the surface didn't change. Any ideas?
> >
> > 2)
> >
> > The reason I am editing wm.mgz is to try to include some cortex that was
> > left out of the pial surface. I don't suppose there's any way to fix
> > this without editing wm.mgz? Because the white matter surface actually
> > looks fine the way it is.
> >
> > Thanks,
> > Eli
> >
> > Houston Methodist. Leading Medicine.
> >