Hi,
In general we do not recommend using non-linear registration between
a functional image (EPI) and a structural image. This is especially true
if you are having trouble getting the initial rigid-body registration to work
well. Can you describe in more detail what "results are not good" means?
Is it just that they are poor in the areas that are distorted? Or are they
off globally (e.g. shifted up or down)? Or are they completely and
utterly wrong? Finally, do you have any fieldmaps that can be used to
correct for the distortions?
All the best,
Mark
On 14 Jan 2011, at 13:16, stefania scopelliti wrote:
> Dear all,
> I tried to register my functional data to T1 data by FLIRT. I perfomed different kind of registration (rigid, affine...) but my results are not good. So I tried with FNIRT (with default option) but my result is really smooth and distorted.
> Can anyone explain improve my FNIRT transformation? Is it possible that the problem could be related with my EPI data, probably too much distorted?
>
>
> Thank in advance
> Stefania
>
|