Just to clarify, "at each point" means at each same-numbered node in the mesh for the white and pial surfaces. (Right?)

What has never been clear to me is, in what sense are the same- numbered nodes in those two meshes "the same point"? Is this accepted as a convenient convention based on the fact that the algorithm grows the pial surface out from the white surface, or is there something specific in the algorithm to assure that a node ends up in "the same place on the surface" in some sense that is anatomically meaningful?

On Jul 10, 2008, at 12:11 PM, Bruce Fischl wrote:

Hi Ryan,

1. Yes, it the average of the shortest distance between white and

pial, and between pial and white at each point.

2. Yes, in part although other things go into it. The wm.mgz is a

starting point (described in many technical papers).

3. The control points will only fix some issues. If the ?.orig

doesn't follow the filled.mgz volume then there is a topological defect this is being fixed incorrectly, and you'll need to edit the wm.mgz to correct it (I think there is a tutorial on this).

cheers, Bruce On Thu, 10 Jul 2008, Ryan Scotton wrote:

...

Hi all,

I am fairly new to FreeSurfer and I've just finished initial runs on about 140 MRI images. Now, my job is to go through and do a little QC that helps ensure that the cortical thickness values are as accurate as possible. Now, I have a few questions:

1. Am I right in assuming that cortical thickness values are

calculated as a function of the distance between the white and pial surfaces?

2. How exactly are the white and pial surfaces calculated? Do

they utilize the wm segmentation?

3. Since I thought this was the case, I have tried cleaning up a

couple images by using control points (using case 4 in the tutorial as a guide), hoping that the white and pial surfaces would be redrawn in a more accurate way. However, the problem remains. The surfaces are not including parts that are obviously white matter and gray matter.

So basically, I'm thinking I'm missing something on the best way to fix white and pial surfaces when they're not accurate enough. Could anyone offer me some advice/a link to a site which explains the best way to handle such problems?

Thanks, and sorry for the looooong post.

Ryan

---

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

-- -dave---------------------------------------------------------------- "Pseudo-colored pictures of a person's brain lighting up are undoubtedly more persuasive than a pattern of squiggles produced by a polygraph. That could be a big problem if the goal is to get to the truth." -Dr. Steven Hyman, Harvard