I wanted to create a Freesurfer surface on some of the aseg ROIs.  I was able to run mri_tesselate, mris_smooth, mris_inflate and mris_sphere some of the subcortical volumes for creating a surface.  

The mri_sphere had a big chunk that was deleted and so I wanted to run mris_fix_topology to fix this issue.  In trying to run this command it appears that mris_fix_topology tries to correct based off of a brain sphere (which this ROI is clearly not) and therefore it fails.  I plan to then run subsequent steps such as mris_euler number and mris_remove_intersections  

1) Is there any way to fix the topology of this ROI using freesurfer tools or a way to work around this issue?  Do you feel the sphere is needed for a small subcortical surface (ie globus pallidus) or is this really more for complex structures such as cortical surfaces?   

2) Is there a recommended set of steps to accomplish creating a surface from the ROI that is different from what I stated?

Thanks,

Ajay