Hello,

I'm stuck and not quite sure what is going on.

Unfortunately, I do not have the original dicoms, therefore I am converting Analyze formatted data to .nii using an outside program. When I use 3dinfo on the newly created .nii data it is oriented as LPI, however, when I view the data in the AFNI viewer axial looks good (eyes forward, back of head in back), but both sagittal and coronal have the brain as being upside down. I first tried using 3dresample to change the orientation to a new orientation (e.g. LPS) and while the header information changed (e.g. LPS), when I loaded the data up into the viewer the data is unchanged - both sagittal and coronal are upside down.

Then back on the original .nii data I tried "3drefit -orient LPI" and while this did not change the header information (3dinfo still showed LPI), now when I load the data into the viewer everything is upright and looks correct. So it looks like my conversion process to .nii is not setting the header orientation in a way that the AFNI viewer can read it, so I need to use 3drefit to correctly set the header.

My question is - why does the header indicate LPI but the viewer does not seem to interpret this correctly? If the AFNI viewer cannot interpret the orientation information correctly, can I trust that 3dinfo is correctly reporting the data is LPI to begin with - i.e. is the data I am converting to .nii really LPI? Any help would be greatly appreciated!


Thanks,
David