Dear Freesurfer experts,

I have recently processed a group of subjects  and am interested in the eTIV and brain volume to estimate the brain parenchymal fraction.  The values that eTIV from the asegstats seem to be underestimated.  

I understand that the eTIV is based on an atlas scaling factor determined from determinant of the affine transform to the Talairach  atlas.  Is the transform to Talairach space completed after the skull has been removed? If CSF surround the brain is missing  from the brainmask, would this result in a smaller estimate of total intracranial volume?  Our subject population is older with the presence of atrophy.  Would you recommend manual editing of the brainmask and reprocessing to include CSF surrounding the brain to get a better representation of the intracranial volume?

Thanks for your advice and any clarification of the method can offer.

Regards,

Cheryl