> For fMRI data, SPM8 still relies on the *.mat files. The reason for this is
> that the nifti file can only have 1 transformation matrix in it (I learned
> this about a month ago). If you remove the *.mat files from a directory and
> display the images, you will notice that they shift. The classic example is
> motion correction. After you motion correct the data, you have a mat file.
> If you remove the mat file, you undo the motion correction.
This only applies to 4D NIfTI files. If the fMRI are a series of 3D
volumes, then there should be no .mat files.
When using old ANALYZE format images, there is always the chance of
some left-right confusion about image orientations. It may be an idea
to check the files using the spm_orientations function. This will
show you where SPM considers the first voxel in the image to be. Of
course, you would need to know the true orientations in order to say
whether SPM is treating the orientations correctly.
Best regards,
-John
|