then you didn't fix it properly. Did you find the source of the defect? This is kind of an acquired skill as you have to page through a number of slices almost by definition, and sometimes need to look in different orientations. If you point us at the data (and the voxel coords of the problem) we'll take a look.

cheers, Bruce

On Mon, 22 Sep 2008 dahlia@nmr.mgh.harvard.edu wrote:

Thanks Bruce. But what happens when I've already done the manual edit, run autorecon2-wm and still see the defect?

...

Hi Dahlia, this usually means there was a topological defect that was fixed incorrectly. You'll need to edit the wm.mgz to remove it, then run autorecon2-wm.

cheers, Bruce On Mon, 22 Sep 2008 dahlia@nmr.mgh.harvard.edu wrote:

...

Hi,

What should I do when the white volume looks right (e.g. filled in where needed) but the white surface doesn't follow the boundary of the volume? I've done recon-all with autorecon2-wm and autorecon3.

Thanks! Dahlia. _______________________________________________ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer

Hi Dahlia,

this is because there are voxels labeled cortex through which the lh.orig surface passes. If you changed those to accumbens then the surface wouldn't deform inwards there. I guess this is another case that should be handled by the stuff that detects non-cortical regions and freezes the surface, but because of the gm voxels it doesn't.

cheers, Bruce

On Tue, 23 Sep 2008 dahlia@nmr.mgh.harvard.edu wrote:

I'd appreciate that. For example: data is at ~dahlia/structurals/subj30/mri/wm.mgz and the coords are [133 126 155]. [in this case the corrections were made automatically, not manual edits - but the white surface doesn't follow the volume edits) Thanks!

...

then you didn't fix it properly. Did you find the source of the defect? This is kind of an acquired skill as you have to page through a number of slices almost by definition, and sometimes need to look in different orientations. If you point us at the data (and the voxel coords of the problem) we'll take a look.

cheers, Bruce

On Mon, 22 Sep 2008 dahlia@nmr.mgh.harvard.edu wrote:

...

Thanks Bruce. But what happens when I've already done the manual edit, run autorecon2-wm and still see the defect?

...

Hi Dahlia, this usually means there was a topological defect that was fixed incorrectly. You'll need to edit the wm.mgz to remove it, then run autorecon2-wm.

cheers, Bruce On Mon, 22 Sep 2008 dahlia@nmr.mgh.harvard.edu wrote:

...

Hi,

What should I do when the white volume looks right (e.g. filled in where needed) but the white surface doesn't follow the boundary of the volume? I've done recon-all with autorecon2-wm and autorecon3.

Thanks! Dahlia. _______________________________________________ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer