Hi,
Despite what fslcreatehd suggests, a nifti image does not explicitly
contain an origin. The 0,0,0 mm point is specified implicitly by the
sform or qform. Hence you should be modifying these with
fslorient rather than creating a new image with fslcreatehd.
I suspect that you have made a radiologically ordered header
with fslcreatehd (negative determinant of the sform or qform
matrix) whereas your origin image data was neurologically
ordered (positive determinant).
It might be worth saying why you need the origin set. Maybe
you actually need an Analyze image rather than a nifti one.
If you do need nifti, you can set the sform or qform matrix
explicitly to do whatever you want, but it can be quite a
technical exercise.
All the best,
Mark
On 17 Nov 2009, at 15:26, Sebastian Rodrigo wrote:
> Dear FSL experts,
>
> I want to modifiy the origin of 4D nifti EPI images using fslcreathd :
> ex : fslcreatehd 64 64 34 201 3.75 3.75 3.80 5 31 41 14 4 test
> to set origin to 31 41 14 (64², 34 slices, 201 TR set to 5 sec)
> An unexpected flip along X was obtained. Some options missed ?
>
> Best regards
>
>
> Sebastian
>
|