Dear Philipp,
I do not believe anyone has done the specific study you are talking
about, but we have a lot of experience with registration and I would
strongly argue for using a T1-weighted image as an intermediate,
regardless of whether you have fieldmaps or not. If you do have
fieldmaps then that is better, but without them the registration is
still a *lot* better conditioned by doing EPI-T1 and then T1-standard
as the anatomy is shared in EPI-T1 and a lower DOF registration
can be done, which is more robust and copes better with the
distortions and signal loss in EPI, whereas going straight from EPI
to the standard often induces unreasonable scalings and skews
based on the distortions and areas of signal loss in the EPI.
All the best,
Mark
On 5 Mar 2011, at 12:15, Saemann Philipp wrote:
> Dear FSLers,
>
> I have a question on T1-supported registration of EPI time series for MELODIC:
>
> Is this step (i. a. adding a structural image (after BET) for
> coreg/normalisation purposes) of advantage in terms of group reg quality,
> if the EPI images cannot be undistorted (FUGUE) before, or is undistortion
> an essential step?
>
> (Unfort. fieldmaps are not available)
>
> Has anybody compared pure EPI based (linear) registration with (a)
> T1-supported linear or (b) T1-supported linear/nonlinear registration
> in the context of MELODIC?
>
> vG Philipp
>
>
> Max Planck Institute of Psychiatry
> NMR Research Group
> Kraepelinstr. 2-10
> 80804 Munich
> Mail: [log in to unmask]
> Phone: 0049-89-30622-413
>
|