Hi,
The processing of your fieldmap looks absolutely fine.
Make sure you also put in the skull stripped magnitude image into FEAT,
as eliminating the edge voxels can be important.
I'm afraid that we don't have an easy way to put in the nosearch option
in the GUI. What you could do instead is edit featlib.tcl to put this
into
the appropriate flirt call, which should be:
fsl:exec "${FSLDIR}/bin/flirt -in EF_D_example_func -ref
FM_D_fmap_mag_brain_siglossed -omat EF_2_FM.mat -o grot -dof 6 -
refweight FM_D_fmap_sigloss"
Just add the -nosearch option to this line and that should do it.
All the best,
Mark
On 17 Oct 2008, at 19:04, Kyle Kern wrote:
> Hi,
> I am trying to unwarp some fMRI data from a Siemens Allegra scanner
> using a gre_field_mapping sequence and need to prepare the field map.
> The intensity values for the reconstructed phase image range from 0
> to 4094, and the TEs were 4.88 ms and 7.34 ms, so from what I gather
> from other posts, I should do: fslmaths [phase_input] -sub 2047 -mul
> 3.14159 -div 2047 -div .00246 [radians_per_sec_output] such that
> the final
> range is -1277 to 1277 (and mask with the skull stripped magnitude
> image). And I should be able to input this into fugue right?
>
> Also, is there a way to adjust the registration in the b0 unwarp
> option in FEAT so that it does no search? My data has very few
> slices in the z
> dimension, but all my data is practically aligned already (fieldmap,
> epi, matched bandwidth and a structural scan). When I use the b0
> unwarp
> option, it reorients the data in the wrong direction when matching
> the fieldmap to the epi. Using 'no search' in the registration step
> always works
> to align the epi to the matched bandwidth to the hi res structural,
> so I assume it would be helpful if I could include no search for the
> fieldmap
> somehow as well.
>
> thanks a lot,
>
> Kyle Kern
>
|