Hi,
fslorient will allow you to change the qform and sform to be whatever
you want.
I think that this is probably what you are after. You will, however,
have to figure
out the right way to change the matrices. Easiest way is to work out
which direction
is A-P in voxel coordinates (look at FSLView) and then modify this
part of the
matrix appropriately. Note that you will need to set the sform and/
or qform *code*
to be non-zero first, if they are currently set to zero (check this
with fslhd).
Also note that fslorient will keep the labels attached to the same
parts of the
brain and will not do what you want if either the sform or qform
codes are non-zero.
Only when these codes are zero will the swapping of the data order
also change
the labelling (as the labels will be based on an unchanging default
wrt voxel coordinates
in that case, but otherwise it will modify the sform/qform to also
move these labels).
Hope this helps.
All the best,
Mark
On 24 Jan 2008, at 02:28, Dianne Patterson wrote:
> Take a look at fslswapdim...it will allow you to change the
> orientation any way you want...it just takes some playing around.
>
> -Dianne
>
> On Jan 23, 2008 7:14 PM, Yi-Shin Sheu < [log in to unmask]>
> wrote:
> Kim:
>
> I am not sure how exactly to do it with fsl, but maybe you can try
> some header editing command to do it.
>
> If you have AFNI in your computer, you can use 3drefit to edit the
> header to the right orientation. --> afni is rich in various commands.
>
> I just notice that dcm2niiGUI in MRICRON has a function to modify
> nifti file, it looks like you just need to drag your old nifti file
> to the program and output as compressed FSL (4D NIFTI), it should
> work that way. I do not have data to try it right now, but it
> looks like it worth a try.
>
> Yi-Shin
>
>
> On 1/23/08, Kimberly Chiew <[log in to unmask]> wrote: Thanks
> Yi-Shin, I only see information on changing L-R orientation of the
> images with fslorient. Do you know if it's possible to change the A-
> P orientation?
>
> Kim
>
>
> On Jan 23, 2008 4:56 PM, Yi-Shin Sheu < [log in to unmask]>
> wrote:
> Hello:
>
> Method 1: Use MRIConvert or MRICRON's dcm2niigui to convert your
> data from dicom to nifti.
> Method 2: After fslswapdim, use fslorient to change the label of
> the header.
>
> Yi-Shin
>
>
> On 1/23/08, Kimberly Chiew <[log in to unmask] > wrote: Hi
> there,
> I have some data that was preprocessed with other imaging software
> that I have converted into NIFTI format to work with in FSL. When I
> view the data in FSLview, it's clear that the image is incorrectly
> oriented on the y-axis (the back of the brain is labelled anterior,
> and the front of the brain is labelled posterior). I have tried to
> play around with the fslswapdim (avwswapdim) tool to reorient this,
> but it flips the brain as well as the axes leaving the anterior and
> posterior labelled incorrectly. Is there any way to correct for this?
>
> Thanks,
> Kim
>
>
>
>
> --
> Sheu, Yi-Shin
>
>
>
>
> --
> Sheu, Yi-Shin
>
>
>
> --
> Dianne Patterson, Ph.D.
> [log in to unmask]
> ERP Lab
> University of Arizona
> 621-3256 (Office)
|