Alignment problem

[swallowwly]

Hi AFNI team,

I had trouble aligning anatomy with vr_base. The following is my code:

align_epi_anat.py -anat2epi -anat FreqAP05_MPRAGE_c1234_U+orig
-save_skullstrip -suffix _al_junk
-epi vr_base+orig -epi_base 0
-epi_strip 3dAutomask
-ginormous_move
-volreg off -tshift off

The alignment result (*al_junk) is usually that the brain is rotated off by some angles and mismatch the anatomy brain (see attached).

I’m using Unifized data. I’ve tried different ways including: skull-stripping before aligning, different cost functions (lpa, lpc/+zz and others), other skull-strip method, various movement level (big_move, ginormous_move etc.) and none of these could align well (some was able to improve it to some extent).

I never had this problem before with other subjects, so not sure why this one would happen.

Can you please help with this? Thanks!

-Lingyan

[rickr]

Hi Lingyan,

What are you showing here, is it the “aligned” anat on top of the original anat? If so, those are not expect to be aligned. What does an image of vr_base overlayed on the aligned anat look like?

Since you seem to be doing this (at least originally) via afni_proc.py, have you looked at the alignment results in HTML QC report?

• rick