Bruce, Doug, Thomas, thank you so much, things are getting slowly clearer now
I can try to explain better what i want to do, so that perhaps it will be easier for you too to tell what is the best way to approach this problem
essentially we have a number of subjects, with already extracted white matter fiber tracts (in .trk format) The idea was to transform these fibers back to freesurfer space (is it the talairach space? in any case, is this what happens by using the talairach.xfm matrix?), and then compare them in this space, as, i would guess, there is still some sort of deviation from the standard fiber pathway.
Is this unreasonable?
also, just for visualization purposes, i was thinking of rendering these fibers in their context, but the average brain is, as you point out, a bit too blurry. Would it make sense here to bring something like MNI152 into the freesurfer space, and use this one (and its segmentation) instead of the fsaverage as an illustrative atlas in freesurfer space?
Paolo
On Mon, 7 Nov 2011 11:41:46 -0500 (EST) Bruce Fischl fischl@nmr.mgh.harvard.edu wrote:
Hi Paolo,
Are you showing individual subject data in the MNI space, or group average data? In general showing template brains and average data is not a good idea. It misrepresents the spatial localization accuracy of the data. You should try to use averages of the anatomy for groups of tracts of average fmri data or other group stats/data.
cheers Bruce
On Mon, 7 Nov 2011, Paolo wrote:
I think it is well possible. And in principle i just wanted to use it for presentation, but the fsaverage doesnt really look like a "template" brain, rather like an average
that's why i was wondering whether there is a more realistic template available in freesurfer space, or how i can create one (the idea was to recon-all the MNI152, which to me resembles much more to a brain)
would it be the correct approach, or do you have other suggestions? by the way is it normal that recon-all on mni152 takes just ages to complete? (>12 hours on an athlon64, dont know exactly how long, i always stopped it before it finished)
Paolo
On 07-Nov-11 16:03, Thomas Yeo wrote:
I am probably wrong, but doesn't the volumetric part of fsaverage correspond to MNI305 space rather than MNI152 space?
--Thomas
On Mon, Nov 7, 2011 at 9:00 AM, Bruce Fischlfischl@nmr.mgh.harvard.edu wrote:
Hi Paolo,
I think fsaverage has this in the aseg.mgz.
cheers Bruce On Mon, 7 Nov 2011, Paolo wrote:
In the meanwhile i did some research, and it looks like freesurfer does not use MNI152 as default template, am i right? it seems that the default template is the fsaverage subject, which in fact does not look like very realistic.
Is there any way to get a realistic brain& segmentation that is "idealized" enough, and still in talairach space (freesurfer space)?
maybe the MNI152 after recon-all and mri_transform to apply back the transformation?
i am sorry if these questions look unreasonable, i am new of freesurfer, and still have problem orienting myself in the documentation :)
best, Paolo
On 05-Nov-11 00:57, Paolo wrote:
Hello
I am a new user of freesurfer. I was wondering whether the brain regions for the MNI152 dataset are available anywhere (i mean the template brain regions for the default MNI152 template).
I have also another question: i have the dti fiber tracts extracted and computed and already in .trk format. I was wondering: to bring them back into the template MNI152 space, is it sufficient to apply the talairach.xfm transformation matrix to the vertex data, or do i also need to use the m3z files (which look like a nonlinear deformation field)?
thanks in advance! Paolo