Hi,
There isn't much to go on here, but you can try the -nosearch option
if they start roughly the same alignment. Also, make sure you check
that the voxel sizes are correct in both images before doing anything.
If your scanner and conversion software has provided useful information
in the sform or qform part of the nifti image, then the -usesqform
option is also good to try.
If there is substantial distortion in your EPI images then you would
benefit greatly from being able to use a fieldmap, assuming you
acquired one.
If none of the above options help then let us know.
All the best,
Mark
On 22 Feb 2011, at 01:28, Van Lee wrote:
> Hi,
>
> I wanna register the EPI image to the T2 weighted one of mouse brain. I use FLIRT with dof = 6 and the cost is set as default. Before that I did the skull stripping of the T2.
>
> The result is very bad as the image is totally misaligned and the slices at the beginning and the end are cut off partially.
>
> Can anyone provide me some suggestion?
>
> Thanks a lot!
>
> Van L
>
|