>Date: Wed, 27 Jun 2001 09:36:25 +0100
>From: Krish Singh <[log in to unmask]>
>Subject: Re: Talairach demon
>Dear Tobias,
>You can use the Talairach Daemon if you use Matthew Brett's mni2tal.m
>transformation script
>(although he now has an all new improved version). You can download it from:
>http://www.mrc-cbu.cam.ac.uk/Imaging/mnispace.html
>Jason Steffener has produced a webpage about converting a block of
>coordinates using
>mni2tal and then submitting them to the Talairach Daemon. The page is here:
>http://rumba.rutgers.edu/~jason/more/tal.html
>All the best,
>Krish
>
>pop01518 wrote:
> > Dear SPM'ers,
> > I followed your discussion about the normalization space. Apparently,
> the space used in SPM 99 is not
> > Talairach, but rather MNI305.
> > I am using the Talairach demon of the University of Texas Health
> Science Center to identify the locations
> > of my result coordinates.
> > Now, if the data in SPM are not normalized to Talairach but MNI305 and,
> as my
> > understanding goes, is reported in MNI305 space, is
> > it valid to rely on a localization tool that is beased on Talairach space ?
> > Thanks for any reply in advance.
> > Tobias
Hi Folks,
I've been following this discussion of the use of MNI vs Talairach
space re UTHSC daemon co-ordinates. Krish stated unequivocably that SPM-99
(or 96) output coordinates should be transformed from MNI to Talairach
space via Matthew Brett's mni2tal.m transformation script before submission
to the daemon.
My UCLA browser said I don't have permission to access the Rutgers website
above, so if anyone knows how I can get there please advise.
Meanwhile, from the comparison image on
http://www.mrc-cbu.cam.ac.uk/Imaging/mnispace.html I find that at axial
slices close to the AC the edge of the brain is about x=+/-59 to 63 in the
Talairach atlas (never more than 63 at any slice), but is in the upper 60s
in MNI-space (as expected from the discussion on the webpage). However,
the daemon edge for those slices is in the upper 60s, looking more like
MNI-space. What gives?
Also, since the daemon documentation says that SP maps were
generated from the MNI-derived ICBM305 MRI template, should I assume that
SP maps are already in MNI space while the BA labels and other daemon
information are assuming Talairach space and should be transformed before
submission to the daemon for optimal localization? I'm a bit confused, and
would appreciate it if you could straighten me out.
Many Thanks,
|