External Email - Use Caution

Hey FreeSurfer Devs!

I’m trying to do a surface reconstruction on a rodent-sized animal (tree shrew). I have a good segmentation of the white matter and have been able to generate reasonable looking orig / smoothwm surfaces. I’m running into an error with mris_make_surfaces: MRISedges(): too many faces: 9578 9599 n=9, m=4, k=2. I am running this command manually:

mris_make_surfaces -noaseg -orig_white lh.smoothwm.nofix -white white.preaparc -noaparc -mgz -T1 brain shrewsurf lh

I also hit this error when running autorecon2: recon-all -autorecon2-wm -subjid shrewsurf -hires -nonormalization2 -nofill -noaseg -nosegmentation -nocortribbon -notessellate -no-isrunning

Some details on the MRIs I’m trying to do this recon on: 0.05mm^3 T2. No T1s (poor contrast on 9.4T). I’m trying to recon on the native resolution: mris_convert T2.nii T2.mgz -cs 0.05. I dummied a T1 by segmenting the grey/white matter, assigned voxels values of 90 and 110, respectively, then blurred the image a little so there is a gradient across the grey matter from white matter to out of brain.

I’ve had no issues running mri_tessellate, mris_extract_main_component, mris_smooth, mris_inflate.

Any advice would be greatly appreciated! Mike