I would counter that CSF may be important to account for in certain
patient or age cohorts, or at least the ratio of GM+WM to CSF. So, I
wouldn't discount it out-of-hand, but Simone raises important points
about measurement error that complicates its inclusion in TIV.
It has been argued that GM+WM might be a better measure to use when the
assumption of TIV relates to normal aging cohorts, as there is a natural
age-related pattern of CSF increase and GM+WM decrease over time.
GM+WM+CSF in individuals in their 30s and 40s is thought to represent
the individual's maximum CNS growth potential, but this is not the
proper correction to use when individuals are in their 50s or beyond
because you are correcting for a past potential and not current CNS
tissue volume. With our elder patient samples, we opt for GM+WM volume
as a covariate or nuisance variable this reason.
Jeff
------------------------------
Jeff Browndyke, Ph.D.
Duke University Medical Center
Dept. of Psychiatry
------------------------------
-----Original Message-----
From: SPM (Statistical Parametric Mapping) [mailto:[log in to unmask]]
On Behalf Of Simone Reinders
Sent: Tuesday, December 16, 2008 6:56 AM
To: [log in to unmask]
Subject: Re: [SPM] total intracranial volume
Dear Suzanne,
Personally, I would recommend not to include CSF in the TIV
calculation in SPM5. Just use GM and WM (and use the mw* files, not
the smoothed files).
The segmentation of CSF in SPM5 is not very precise (see also John his
remark on this in the DARTEL guide).
I have not yet looked at SPM8, but it should be better there due to
the inclusion of more priors.
Good luck,
Simone.
P.S. I have included my little script, which is by far not as fancy as
Ged Ridgway's but might be of some help?
On Mon, Dec 15, 2008 at 10:29 PM, Suzanne Hilleary
<[log in to unmask]> wrote:
> Hello, community.
>
> This question of how to get total intracranial volume has been asked
> numerous times in different variations, but the answer still seems to
be a
> little unclear...
>
> 1. Is it appropriate to get an estimate of TIV by summing GM, WM, and
CSF
> volumes?
> 2. If so, what is the procedure within SPM5 for calcuating GM, WM, and
CSF
> volumes?
>
> Any help would be appreciated, as I am very new to this program.
>
> Best,
>
> Suzanne
>
> --
> Suzanne Hilleary, M.A.
> Clinical Psychology Doctoral Candidate
> Fuller Graduate School of Psychology
>
--
Dr. A.A.T. Simone Reinders, MSc PhD
King's College London
Institute of Psychiatry (IoP)
Box P063, De Crespigny Park
London SE5 8AF
United Kingdom
|