I am analyzing some nifti-format data (anatomical and EPI), and AFNI is giving me the obliquity warning when viewing the images in the GUI. However, the images look perfect. After running 3dWarp -deoblique, the images are rotated off-axis by the amount initially indicated as deviant from plumb (approximately 28 degrees).

I believe these volumes are not actually be oblique, but for some reason, AFNI thinks they are. I initially presumed the GUI was performing warp-on-demand resuling in the proper cardinal orientation on the non-warped images, but after further reading, I understand the GUI probably isn't performing any warp-on-demand at all, and the properly oriented original images are simply being presented as they are.

One possible cause for the problem, I'm thinking, is that we acquired these data using z-shimming in orbitofrontal cortex, which applies an oblique gradient in OFC to counter the expected gradient of the susceptibility artifact, improving the signal in that region. Could this have anything to do with why AFNI is reading obliquity where there is none? Is there an easy way to correct it?

thanks,
George Monteleone