Hi Mike,
the alignment whole head is used to determine the atlas scaling factor,
which itself was computed using some manually generated ICV values. Did
Randy use skull-stripped brains? I didn't remember that and would have
thought that defeated the purpose since you'll strip the sulcal CSF as well
from only a T1 image. I'll have to go look at that paper again (or ask
Randy :>)
Bruce
On Thu, 4 Sep
2008, Michael Harms wrote:
Now I've become confused.
ICV is based on the determinant of talaiarach_with_skull.lta, for which
the skull is not stripped. So, I've always assumed that basically
everything -- wm, gm, csf, cerebellum, dura, AND skull tissue -- all
"contribute" to the definition of the transform (using whatever cost
metric that the lta registration uses).
Is this correct? If so, a question I've had for some time is to what
extent the skull itself (and not the wm, gm, csf, cerebellum) drives the
talaiarach_with_skull.lta transform, since in principle, shouldn't an
intra-cranial volume estimate be determined solely by the skull?
It is also probably worth noting that this approach for ICV estimation
is motivated by Buckner 2004, which actually computed the transform
using a skull-stripped volume. I've always assumed that this is part of
the reason that the FS folks derived their own scaling factor (for
converting the determinant to an actual volume estimate) in a sample of
22 brains, rather than using the scale factor from the Buckner paper,
which had a much larger sample (147 subjects). Notably, the two
different scale factor differ considerably (2150 in FS vs. 1738 in the
Buckner paper). Whether this is due to the use of skull-stripped vs.
non-skull-stripped volumes in the registrations, or rather reflects an
artifact of different populations (and a potentially less generalizable
FS scale factor due to the much smaller N) remains to be established,
correct?
thanks for clarifying,
Mike H.
On Thu, 2008-09-04 at 07:48 -0400, Bruce Fischl wrote:
yes, that's correct.
Bruce
On Wed, 3 Sep 2008, Jeff Sadino wrote:
Hello,
I am measuring the shrinkage of the limbic system over time for several
subjects and need to know what is included in the ICV value in aseg.stats to do
this. We are putting our raw scans into freesurfer without any manual edits
until after QA at the end. My understanding is that the brain gets registered
to the talairach template, the skull gets stripped, and then everything inside
the skull - wm, gm, csf, cerebellum - gets included in the ICV (via scaling of
the determinant of talairach_with_skull.lta)? Can someone confirm this work
flow for us?
Thank you very much,
Jeff
_________________________________________________________________
See how Windows Mobile brings your life together?at home, work, or on the go.
http://clk.atdmt.com/MRT/go/msnnkwxp1020093182mrt/direct/01/
_______________________________________________
Freesurfer mailing list
Freesurfer@nmr.mgh.harvard.edu
https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
_______________________________________________
Freesurfer mailing list
Freesurfer@nmr.mgh.harvard.edu
https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
_______________________________________________
Freesurfer mailing list
Freesurfer@nmr.mgh.harvard.edu
https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
