Hi everyone,

I would like to produce lh.white, lh.area, and lh.curv files of the type produced by mris_make_surfaces, but where the input is a mask of the left hemisphere white matter alone.

Specifically, suppose I start with the file lh.wm-mask.nii and use mri_convert to make lh.wm-mask.mgz. I can then follow the -tessellate and -fix steps of the -autorecon2 section on the webpage

http://surfer.nmr.mgh.harvard.edu/fswiki/ReconAllDevTable

to produce the file lh.orig, one of the required inputs to mris_make_surfaces.

The only other files required to have mris_make_surfaces produce lh.white, lh.area, and lh.curv files, seem to be the files filled.mgz, brain.mgz, and wm.mgz. Since wm.mgz is the output of a call to mri_pretess from the -segmentation step of the -autorecon2 section, I can create wm.mgz by using the command

mri_pretess lh.wm-mask.mgz 1 wm.mgz.

My inclination at this point is to create copies of wm.mgz that are given the names filled.mgz and brain.mgz. I can then run

mris_make_surfaces -whiteonly -noaparc -mgz subjectID lh

and the desired files are produced, but I'm not sure what the effect of using wm.mgz as both filled.mgz and brain.mgz would have on the output.

Given this setting, my question is: what are the roles played by filled.mgz an brain.mgz in creating the output of mris_make_surfaces?

Thanks in advance for your help, David Romano