My experience with eTIV is that it's great with a perfect Talairach transform but otherwise is less accurate than other measures. Manually fixing the transforms and reprocessing as necessary will result in great eTIVs but requires quite a bit of manual work.

What we do instead is just use the brainmask volume (but our data are MRIs of non-demented older adults so there's not typically extensive atrophy). We've found that brainmask has a Dice Similarity Coefficient of 0.95 and an ICC of 0.92 with manually traced intracranial masks (inferior termination on a straight line between the lowest portion of the clivus and occipital bone). eTIV had an ICC of 0.67 with our manual masks (but we didn't fix the transformations).

Jared

____

Jared Tanner, Ph.D.

Research Assistant Professor

Clinical and Health Psychology

University of Florida

---------- Forwarded message ----------
From: "Harms, Michael" <mharms@wustl.edu>
To: "freesurfer@nmr.mgh.harvard.edu" <freesurfer@nmr.mgh.harvard.edu>
Cc:
Date: Sun, 21 Feb 2016 18:10:12 +0000
Subject: Re: [Freesurfer] eTIV question

Hi,
Why not use a measurement of brain size rather than “eTIV”?

cheers,
-MH

--
Michael Harms, Ph.D.

-----------------------------------------------------------
Conte Center for the Neuroscience of Mental Disorders
Washington University School of Medicine
Department of Psychiatry, Box 8134
660 South Euclid Ave.Tel: 314-747-6173
St. Louis, MO 63110Email: mharms@wustl.edu