We are new users of freesurfer (linux v 4.0.5) and have the following question:
We have some .4dfp.img MPRAGE files which we are trying to use (from Wash U). When we run recon_all (1st step) or mri_convert on its own, the brain comes up upside down as if it was rotated 180 degrees around the sagital plane (see attached picture of the 4dfp.img file viewed in scuba). Since when we view the original 4dfp.img file or the orig.mgz file they are both in this orientation it appears that this is a problem with the original files.
If we use tkmedit to rotate the brain around the sagital plane once we're in .mgz format the recon steps (including normalization which, obviously, would fail previously when the brain was upside down) would complete fine. However I'd prefer to rotate these brains using mri_convert (or another program) so I can rotate the original files; which is why I'm writing now.
To determine what the current orientation for these images was I opened the original files in scuba & ran MRI convert like this: mri_convert --in_orientation ASL -i vcXXX.4dfp.img --out_orientation ASL -o vcXXX.4dfp.ASL.img
until I found an orientation which matched the original. The orientation which matched the original was ASL. In this step I just converted from one form to the same form to reduce confusion.
Next I ran several versions of the following: mri_convert --in_orientation ASL -i vcXXX.4dfp.img --out_orientation SAR -o vcXXX.4dfp.SAR.img
however no mater how I changed the 'out orientation' when the files were viewed in scuba they all appeared to have the same orientation as the original (ASL or rotated around the sagital plane).
Is there any more documentation on re-orienting in mri_convert or does anyone have any ideas/experience that can help?
Thanks, Margaret Sheridan Kathy Zhang
