Hi,
You want to use both fslswapdim and fslorient.
This is described on the FAQ at:
http://www.fmrib.ox.ac.uk/fslfaq/#general_fslorient
Using both will change both header and data
so that they match again.
All the best,
Mark
On 2 Feb 2009, at 00:23, Michael Rohan wrote:
> Hi,
> I'm preparing data from a scanner (varian) for nifti format and fsl
> analysis. I have the data in the nifti format, correctly labeled in
> qform
> and sform, but it is in neurological storage order. I was hoping to
> use
> fslswapdim to change the file into FSL's preferred axial HFS
> radiological
> format, but on reading the fslswapdim and fslorient notes am not
> sure what
> they'll do.
>
> A point to note here is that the data, while inconveniently ordered,
> is
> correctly labeled. So I need to swap the handedness of the data and
> the
> header information together, since they match.
>
> I get the following information from the FSL documentation:
> 1. fslswapdim will reorder the data and header together UNLESS the
> requested
> action will change the handedness of the data (my case), in which
> case the
> "data will be swapped but the header will not". (this seems to dash
> my hopes).
>
> 2. Other suggestions include using fslswapdim -x y z and fslorient
> -swaporient; however my data (prepared for all scan orientation in
> this
> neurologically ordered scanner) wouldn't be handled by -x y z in all
> cases,
> and the documentation for fslorient -swaporient states only that it
> "swaps
> FSL radio and FSL neuro" and I'm not sure what that means or what it
> will do
> to a general header.
>
> So, can the fslswapdim program be made to modify the data as I want,
> changing the data and the header consistently from neuro to radio
> while
> preserving their correspondence?
>
> Mike Rohan
>
|