Dear Louis,
Thanks for pointing this out.
It is actually a problem with the practical.
In the practical the fmap_mag_brain image is created, but without making one called fmap_mag.
FEAT looks for a pair of images with names in the format:
somename_brain.nii.gz and somename.nii.gz
and takes the latter as the non-brain-extracted version (you specify the former in the GUI).
If it cannot find the latter then it uses the former for both, which is not ideal but can still work in many cases (just not as reliably).
So your understand is good, and you should be giving it both a brain-extracted and non-brain-extracted (i.e. whole head) version.
I will change the practical instructions to fix this.
Thanks very much,
Mark
On 28 Oct 2013, at 09:02, Louis Shue <[log in to unmask]> wrote:
> Dear all,
>
> The script "epi_reg" can be used to perform simultaneous EPI distortion-correction and registration of EPI to structural images. The list of optional parameters includes fmapmag and fmapmagbrain, as noted here.
>
> http://fsl.fmrib.ox.ac.uk/fsl/fslwiki/FLIRT/UserGuide#epi_reg
>
> However, I am puzzled by the log after running the following FSL tutorial (http://fsl.fmrib.ox.ac.uk/fslcourse/lectures/practicals/reg/index.html). When using Feat GUI, with B0-unwarping option selected, it seems that identical values were passed to fmapmag and fmapmagbrainl!?
>
> May I know what is the difference between fmapmag and fmapmagbrain? Along similar lines what was the rationale behind having separate parameters?
>
> Regards,
> Louis.
|