Hi Roman,
The .img part simply contains the intensities of the voxels, the .hdr
part contains the information necessary to decide where in world/mm
coordinates particular voxels should map to. It isn't the case that
the .img is the "proper place" for the header info; SPM5 is doing the
correct thing according to the NIfTI standard.
To see correctly oriented images in software that doesn't understand
the NIfTI headers, it is necessary to resample the image data so that
the headers of the new image can be simpler. Note though that this is
not simply a case of shuffling around some header information, the
voxel intensities must be interpolated in the new image, hence some
blurring may occur.
To do this resampling, I think you can use the reslice option under
realign. Or there is probably something in the Volumes toolbox. Or you
can use this script:
http://www.cs.ucl.ac.uk/staff/gridgway/vbm/resize_img.m
Alternatively, there is a new version of MRIcro that does understand
the NIfTI orientation information, which might be the best solution
for you:
http://www.sph.sc.edu/comd/rorden/mricron/
Hope that helps,
Ged.
Roman Rodionov wrote:
> Dear List,
>
> I've learnt that SPM5 saves info about reorientations of an image to .hdr
> file. As image file saved by SPM5 is a NIFTI file the header is embedded
> into .img file. So MRIcro can't visualise reoriented image unless changes
> made in the .hdr file are not copied into .img file.
>
> I think I'm not the first who encountered with this issue. Could you please
> let me know if there is a script (in SPM5 of the separate one) which copies
> header info from .hdr file into proper place in the .img file?
>
> Best wishes,
> Roman
>
|