One of the many reasons I love mri_reface is that it can accomodate DICOM in and out! BUT, I think I'm running into an issue with registration failure due to an image origin that is too low. In the attached picture, the underlying image is a TSE T2-weighted image, the image origin is marked with the green crosshair, the white outline is the difference between the original image and the _deFaced.nii (essentially where the refaced mask is), the red outline is a quick brainmask using Freesurfer synthstrip, and the green is where the refacing and brainmask overlap. The picture on bottom shows the refacing on top of the head.

Since I really want to stay in DICOM space, is there a way to either:

1. Fix the origin in the input DICOMs before mri_reface?
   OR
   
2. Provide a starter/hint transform to the initial affine registration?
   

Thanks for your help!