Hi Bruce,
Actually I also acquired a 1mm dataset at 3T and 7T, which works fine using FreeSurfer. But I'm trying to compare the results with a 0.6mm (0.5mm later) dataset of the same subjects scanned at 7T.
I tried selecting a seed point for the cutting planes following your advice and using the tutorial in the wiki. Instead of using the seed point I specify mri_fill is looking for another one "in the neighborhood". For example if I choose the coordinates 193 197 174 mri_fill uses 186 220 216, resulting in also removing most of the occipital lobe.
Is it possible to force FreeSurfer using the seed point I choose? Or do I have to do something else so mri_fill recognizes the seed point I choose to be valid?
Thanks in advance, Regards, Falk
-----Original Message----- From: freesurfer-bounces@nmr.mgh.harvard.edu [mailto:freesurfer-bounces@nmr.mgh.harvard.edu] On Behalf Of Bruce Fischl Sent: Thursday, July 30, 2009 2:39 PM To: Falk Lüsebrink Cc: freesurfer@nmr.mgh.harvard.edu Subject: Re: [Freesurfer] Processing data with -noaseg flag
Hi Falk,
can you really not afford the 6 minutes or so of scan time to get a good 1mm dataset? We have analyzed a bunch of 7T data successfully, although it probably isn't as automated as 1.5 and 3T. The error you are getting we only see when you have run out of disk space. It sounds like the surface file is corrupted. As for the cutting planes, you should be able to select a point in the pons above the cerebellar peduncle for the brainstem cutting plane. This should remove the cerebellum, which is required for analysis.
We certainly did many cortical thickness studies before the aseg stuff was integrated into the stream, so it should be possible, it just requires more work.
cheers, Bruce
On Thu, 30 Jul 2009, Falk Lüsebrink wrote:
Hi FreeSurfers,
As the subcortical segmentation fails fairly often on my high resolution data I want to skip it by using the noaseg flag, but Im currently encountering another issue.
1.) During the fix stage of autorecon2 I receive following error
during
mris_euler_numbers (also if I want to view the lh.orig or rh.orig in tksurfer):
rh.orig: face[233615].v[2] = 119048, but face 233615 not in vertex 119048 face list
lh.orig: face[196300].v[1] = 100366, but face 196300 not in vertex 100366 face list
2.) Normally the subcortical segmentation is also used to cut both hemisphere and to remove the cerebellum. The cutting plane between the hemispheres is done very well, but I have never been able to remove the cerebellum. Do I need to remove it manually or are there some ways to remove it automatically? Is it required to be removed?
3.) Im interested in measuring cortical thickness only. But do I need
to
keep something else in mind while not using the subcortical segmentation to process my data?
Thanks in advance,
Regards,
Falk