TIV can be computed by adding the volumes of GM, WM and CSF together.
In older versions of SPM, the CSF was not segmented so well (no
information about skull and scalp in the model). However, the CSF
segmentation generally works better in SPM8 (for most image contrasts -
not necessarily all of them), which should provide more accurate TIV
measures.
The volumes of tissue in the c1, c2 and c3 images can be computed using
the following:
P=spm_select(Inf,'nifti');
N=nifti(P);
for i=1:numel(N)
dat = N(i).dat(:,:,:);
dat = dat(isfinite(dat));
voxvol = abs(det(N(i).mat(1:3,1:3)))/100^3;
fprintf('%g litres\n', sum(dat(:)).*voxvol);
end
Best regards,
-John
To anyone who can help,
We are attempting to do a VBM analysis on SPM8 and would like to
determine the total intracranial volume (tiv). We tried using a
code for SPM5 NewCalc_tivs, however, have been unsuccessful.
Should this code work for SPM8 or is there another code or means
of determining the tiv?
Thanks for your assistance
Phil Gerretsen
________________________________________________________________________
--
John Ashburner <[log in to unmask]>
|