I'm wondering about how to best approach two similar problems with pial surfaces. In one case I'd like to extend it and in the other I'd like to reduce it (for lack of a better term) These first two images are from a brain in which I'd like to extend the surface.
http://i.imgur.com/ziDQd.png http://i.imgur.com/i0sX3.png
In this case focus on the bottom left and right corners of the cortex. You can see, especially in the second image (with aseg.mgz loaded), the pial surface cuts off part of the cortex on both sides. Interestingly, the pial surface seems to adhere to the aparc.mgz file, which is loaded and pretty clear in the first image, where the left out parts of the cortex are just labelled as 'none'.
http://i.imgur.com/aUKiG.png http://i.imgur.com/4BRwv.png
In the second case I want to edit the pial surface so it follows that gyrus toward the midline in the left hemisphere. In this case I've tried control points, specifically along the gyrus, and autorecon2-cp. I also tried to edit the parcellation along the gyrus for several slides followed by autorecon2 & 3. This didn't help anything, so I'm a bit stuck. You can see this brain follows the same trend as the last in which the pial surface follows the aparc.mgz, and the aseg.mgz seems more or less correct.
Thanks and any help would be awesome!
jon
Hi Jon
usually this happens because the wm in the wm.mgz doesn't extend out far enough. Is that the case here? If not, you can also try direct repositioning in freeview under tools->reposition surface.
cheers Bruce
p.s. we also have some newer tools that I've been working on that can use a flair or T2 image for this purpose, if you happen to have a high res one
On Wed, 17 Oct 2012, Jonathan Holt wrote:
I'm wondering about how to best approach two similar problems with pial surfaces. In one case I'd like to extend it and in the other I'd like to reduce it (for lack of a better term)These first two images are from a brain in which I'd like to extend the surface.
http://i.imgur.com/ziDQd.png http://i.imgur.com/i0sX3.png
In this case focus on the bottom left and right corners of the cortex. You can see, especially in the second image (with aseg.mgz loaded), the pial surface cuts off part of the cortex on both sides. Interestingly, the pial surface seems to adhere to the aparc.mgz file, which is loaded and pretty clear in the first image, where the left out parts of the cortex are just labelled as 'none'.
http://i.imgur.com/aUKiG.png http://i.imgur.com/4BRwv.png
In the second case I want to edit the pial surface so it follows that gyrus toward the midline in the left hemisphere. In this case I've tried control points, specifically along the gyrus, and autorecon2-cp. I also tried to edit the parcellation along the gyrus for several slides followed by autorecon2 & 3. This didn't help anything, so I'm a bit stuck. You can see this brain follows the same trend as the last in which the pial surface follows the aparc.mgz, and the aseg.mgz seems more or less correct.
Thanks and any help would be awesome!
jon
And in your second set of images, you probably need to edit the wm.mgz. I suspect that there is a topology problem that is messing up the wm.orig
cheers, -MH
freesurfer@nmr.mgh.harvard.edu
-
Bruce Fischl -
Jonathan Holt -
Michael Harms