Hi all,

I've been trying to align low resolution EPIs from one site to partial volume high resolution EPIs from another site. All my attempts have produced what appear to be decent alignments (at least regarding edges), but for one thing: the midline seems to have shifted left. So, the right hemisphere seems to bleed into the left hemisphere, so all the activations from the subsequent analyses are shifted by a neat amount. I considered that perhaps the orientation of the first session's EPIs were perhaps left to right instead of right to left, as it seemed slightly consistent with the anatomical images and the slight skew around the occipital lobe, but this is not consistent with the activations we got from the GLM. Could this have anything to do with the obliqueness warning (underlay/overlay pairs may seem misaligned because of their oblique angle differences)? Lastly, the EPIs do seem as if they've been slightly more clockwise rotated relative to the anatomical...

My procedure:

- S1 anatomical aligned to some S1 EPI (align_epi_anat.py);
- S2 anatomical aligned to some S2 EPI (align_epi_anat.py);
- S1 anatomical aligned to S2 anatomical (3dvolreg);
- S1 EPIs resampled to S2 EPI (for grid mismatch, I think);
- S1 EPIs rotated using earlier 3dvolreg alignment info.

I'm not sure where to upload pictures/data if needed.

Any help much appreciated.

Cheers,
Mahen