Dear Bas
Thanks for the suggestion.
The problem is that we do not normalize our data. In addition ,we use
the functional data to inform MEG analysis for the same subjects and so
need the anatomy to stay intact.
asaf
Bas Neggers wrote:
> Dear Asaf, list,
>
> perhaps you could do it the other way around? E.g., coregister your T1
> to your functional data and leave your functional data where it is?
> When you reslice your T1 during coregistration (if you really want
> that) the T1 would then assume to space of your functional data.
> That's how we usually do it.
>
> It all depends a little on what you want to achieve:
> when you do subsequent normalization, it really does not matter which
> way you take, because images are commonly resliced after normalization
> anyway (and you can set the desired reslice resolution in the
> defaults). It probably takes a lot less time when you coregister the
> T1 to the functional data (without reslicing) and then use that T1 to
> normalize to the MNI template, using the functionals as 'other' images
> that simply hitch the ride to MNI.... and it would save you one
> reslicing step, leading to a bit less image deterioration.
>
> just my 2 eurocents..
>
> cheers,
>
> Bas
>
> Asaf Bachrach schreef:
>> Dear SPMers,
>> We tried to coregister EPI data (3*3*3) with the individual subject's
>> T1 image (1*1*1) with the T1 image as the target. SPM2 automatically
>> resliced the EPI data into 1*1*1 voxels. How can we do this kind of
>> coregistration without changing the resolution of the EPI images
>> thanks,
>> asaf
>
>
|