Dear Freesurfer experts,
I'm converting a batch of nifti files to start a FS pipeline. There is
1 structural image per participant.
The initial 'reconall -i' step does not give any errors.
But, looking at the 'mri/orig/001.mgz' I notice for 1 participant what
appears to be an imperfect conversion.
Hi Bruce
Thanks, it worked. A follow-up question:
I am performing multivariate analysis of the jacobian to cluser
co-variating cortical patches which shrink/expand together. So, is it
correct that if I omit/decrease the jacobian penalty term in the
spherical registration step then I will get more
Hi Alex.
Take a look at the wiki page:
http://surfer.nmr.mgh.harvard.edu/fswiki/FsTutorial/LongitudinalTutorial
you need to create a .Qdecrc file in the qdec directory to tell qdec about
the filenames with the .long in it:
Also QDEC will not know about our new files (e.g. lh.long.thickness-spc
Hi Uri,
is there some other problem? I suspect that if you displayed the nifti in
tkmedit it would look the same. We center the coords and reslice to coronal
to display things in a standard location/orientation.
cheers
Bruce
On Sat, 10 Nov 2012, Uri Hasson wrote:
> Dear Freesurfer experts,
Hi Sourena
yes, that is true. If you are doing a jacobian analysis you probably want
to either turn the metric preservation term all the way up (or just
terminate the warp after the rigid rotational alignment) and analyze the
curvature, or all the way down so that everything is in the jacobian.
