Dear AFNI experts,

I have a question about how AFNI handles oblique functional images.

I decided to re-analyze some data from a friend who uses SPM for preprocessing. Thus I have an anatomical volume (nifti format, plump) and several functional beta estimate volumes (.hdr & .img, oblique).

In MRIcron, the overlay of the functionals on the anatomical is perfect.
In the AFNI viewer, with the images converted to AFNI format, the images are not aligned. Fixing a potential wrong origin [1] with 3drefit -oblique_origin does not fix it, neither does or 3drefit -deoblique does not change anything (also, ORIGIN and last column of IJK_TO_DICOM_REAL show the same numbers). When I apply 3dwarp -oblique2card to the functionals, however, the images are well aligned; similarly, I can use 3dwarp -oblique_parent to resample the anatomical so that it is aligned with the functionals.

So I wonder why this is the case. I assumed that AFNI always does warp-on-demand. Why are the original images not in alignment, but does resampling them with 3dwarp fix the issue?

thanks a lot,

Nick

[1] [afni.nimh.nih.gov]