> When I perform realign and unwrap, I also select the reslice (unwrap)
> option. However, I realized that the output volumes are only affixed u* not
> an ru*. And, when I would like to view the output volumes with MRIcro, it
> gives a warning : " This image includes rotation transformation. You should
> reslice this data before attempting overlay..."??? As a result I get the
> error "images don't all have same orientation & voxel size " when I use
> spm_read_vols to read these images. Am I missing something? Or is there
> another reslicing I need to perform?
Which version of SPM? If you are using SPM5, then use MRIcron rather than
MRIcro (uses NIfTI format - see my previous email to the list).
To see which images cause the problem, then try:
V=spm_vol(spm_select(Inf,'image'));
spm_check_orientations(V);
>
> My second question is that when I do realign and unwrap, I lose my first
> axial slice (i.e. first slice of uvol*.nii is zeros, or mostly zeros). The
> effect is worse if I register to first instead of register to mean. Can it
> be because of high axial motion? I am attaching an example vol.nii and
> uvol.nii and the motion paramters figure for those who would like to
> inspect and help (u*.nii is registered to mean). I really would
> appreciate.
See the "Masking" section of Section 2.2.3 of the manual (or Section 3.4.4).
Note that you should probably not change this setting from the default
though.
Because of subject motion, different images are likely to have different
patterns of zeros from where it was not possible to sample data. With
masking enabled, the program searches through the whole time series
looking for voxels which need to be sampled from outside the original
images. Where this occurs, that voxel is set to zero for the whole set
of images (unless the image format can represent NaN, in which case
NaNs are used where possible).
Best regards,
-John
|