Hi Marieke,

we compute the transform (12 parameter) to tal coords, but don't typically apply it, so you are correct, the T1.mgz is *not* in tal coords. From the subject's mri dir you can use

mri_convert -at transforms/talairach.xfm T1.mgz T1.tal.mgz

to transform the T1 into tal coords (and -ait to transform back if you want)

On Mon, 17 Nov 2008, Marieke Langen wrote:

Dear FreeSurfer people, I have got a very basic question. I would like to (or better: have to) do some manual tracing on my data. For this, I want my brains to be motion- and nu-corrected and in Talairach. Therefore, I apply the -autorecon1 step to my dataset.

I always assumed that the T1.mgz output is in Talairach space, but 1) when I compare it to the original scan and 2) when I look at the processing flow in the ReconAllDevTable, I doubt this. Do I still need to apply the talairach.xfm to the T1.mgz or something like that?

1. Could you please let me know whether the T1.mgz is indeed the scan I need

(thus: is T1.mgz the result of all the autorecon1-steps, accept for skullstripping)? 2) If not: how do I process my scans so that they are motion- and nu- corrected and in Talairach space?

THANK YOU!

Marieke