Hi,
By "remain in its original spacing", do you mean have the same voxel size?
If so then the easiest thing to do is to do step 2 and then use ApplyXFM to create a resampled version with whatever FOV and voxelsize you want. Note that the problem you had was probably due to the FOV not being appropriate. Also note that we recommend using the T1 as the reference, and the EPI as the input. In fact, we recommend BBR for this as it gives _much_ better registrations. Have a look at the FSL course practical on registration, as I think you'll find this quite helpful.
All the best,
Mark
On 25 Jan 2013, at 22:14, Vincent Koppelmans <[log in to unmask]> wrote:
> Dear FSL experts,
>
> I want to co-register a T1 file to a 4D epi file. (The epi files are already co-registered). The problem is that the T1 should remain in its original spacing.
>
> I tried the following:
>
> 1) create an empty file with a the following header that matches the header of the T1 image:
> fslcreatehd 160 256 256 1 1 0.94 0.94 1 0 0 0 16 header.nii.gz
>
> 2) calculate the rigid transformation:
> flirt -dof 6 -in T1.nii.gz -ref EPI.nii.gz -omat rigid_trans.mat -out
>
> 3) apply the transformation with the header file as reference:
> flirt -in T1.nii.gz -ref header.nii.gz -applyxfm -init rigid_trans.mat -out flirt_T1.nii.gz
>
> I tried 3) both with and without "-applyxfm". Both outcomes were not what I hoped for.
>
> Could somebody please point me in the right direction?
>
> Thanks, Vincent
|