Dear Roberta,
If your image from SPM is in MNI space (and I assume it is) then what you are doing is probably fine. However, it might be safer to do the following:
flirt -in VBMimage -ref FIRSTimage -usesqform -applyxfm -out VBM_highres
This will make sure that the coordinate centres are perfectly aligned in the MNI space, in case there are any subtle things like half-voxel shifts due to different coordinate conventions between SPM and FSL. It may give you exactly the same output as what you've already done, in which case they're both fine, but if there is a difference I would go with the result of the above flirt call, assuming that it looks correct by eye.
All the best,
Mark
On 23 Apr 2013, at 11:13, Roberta Vasta <[log in to unmask]> wrote:
> Hi all,
> I need to superimpose VBM (SPM) to vertex analysis results in order to highlight that the latter is more sensitive than the former to subregional changes. I focused on hippocampus.
> VBM results are represented by a 91x109x91 image, while the basestruct built by first3Dviewer (FSL5.0) has dimension 182x218x182.
> I tried to transform the VBM image into vertex one applying FLIRT transformation:
>
> input volume: VBM image
> Output size-Base on voxel dimensions- num voxel: 182x218x182 and voxel size 1x1x1 (vertex image dimension)
>
> Then I superimposed vertex analysis image and VBM converted one by fslview.
>
>
> How can I be sure that the transformation was successful and that the two images are centered at the same point?
> Is there a more rigorous way to transform and then superimpose them?
>
> Thanks
|