stable-pub-v4.0.5 on Suse 11.

I've used the output of FreeSurfer, specifically wmparc.mgz to create a 32-bit volume to which I've applied edits. The two commands I used were:

mri_vol2vol --mov mri/wmparc.mgz --targ mri/rawavg.mgz --o wmparc-in-native.mgz --regheader --interp nearest

mri_convert --in_type mgz --out_type analyze wmparc-in-native.mgz wmparc_in_native_analyze

(The new Analzye set was renamed FSParcel.hdr/img)

After making the appropriate edits, I wanted to put this file back into mgz format/freesurfer space so I ran:

mri_convert --in_type analyze --out_type mgz FSParcel 3065_FSParcel.mgz

mri_vol2vol --mov FSParcel.mgz --targ mri/wmparc.mgz --o FSParcel_FSorient.mgz --regheader --interp nearest

(I assumed that a target of wmparc would be most appropriate sinc that's what I started with but I've tried aseg as well)

I've been playing around with Slicer and trying load the file as a label file to which I create (I believe) vtk models. If I convert wmparc to an Analyze set and then back to FreeSurfer mgz without any edits, I get an exact copy, as would be expected. If I make any alterations to the file and follow my steps, Slicer gripes that it can't parse the header. No matter what dims I use, I get pixels scattered across the screen and it appears that the volume is scrunched to 1 slice. If I run tkregister (as mri_vol2vol suggests)

tkregister2 --mov FSParcel_FSorient.mgz --targ $SUBJECTS_DIR/sub1/mri/aseg.mgz --reg FSParcel_FSorient.mgz.reg

everything looks fine. I find the Slicer documentation a bit lacking and don't know who to turn to. Is there something inherently wrong with my commands or is there a way to write header params to the mgz file so that Slicer will just work?

-Joel