Dear Freesurfer Experts,

I have recently processed some data with the following 2 steps:

1) Perform skullstripping outside freesurfer to generate a nifti file (mostly using AFNI).

2) Run that through freesurfer using the -noskullstrip flag.

Comparing the results from this with just running freesurfer alone, I found that the eTIV was significantly smaller, and not terribly well correlated (R^2 of .65).  When looking at cortical volume measurements, it is generally considered a good idea to correct for eTIV.  How would you recommend doing so in this case?

Freesurfer version 5.3.

Thanks,

-Eric