Dear Tae,
Further to Simon's comments, I think that a standard orientation is
important for any kind of volumetry (but nonlinear spatial normalisation
to MNI/ICBM152 is probably a bad idea, see ref by Simon & Ashburner J et
al. 1997).
We reorient datasets to ACPC with fixed protocols (see papers for
details); affine registering should work too in principle but I have no
experience with it. Whatever you do, very importantly, choose the best
available interpolation strategy for resampling - a high degree B spline
or a widely windowed sinc -, and do stay in 16 bit - this makes a real
difference for tricky structures.
For normalisation issues, have a look at the excellent Free SL et al.
1995 AJNR paper.
All the best,
A
-----Original Message-----
From: SPM (Statistical Parametric Mapping) [mailto:[log in to unmask]]
On Behalf Of Woo Suk, Tae
Sent: 01 February 2007 08:27
To: [log in to unmask]
Subject: [SPM] MRI volumetry with spatially normalized images
Dear SPMers.
I am planning to measure the volumes of frontal and temporal gyri.
And I think that each gyral volumes should be corrected by the volumes
of whole brains or intracranial cavities.
What I want to know is that the gyral volume measured with spatially
normalized MRI have a same effect compared with
the volume normalization by whole brain or intracranial cavities?
And is it more appropriate measuring with affine registered MRI without
warping?
If anyone had experience in this kinds of measurement, any comment would
be appreciated.
Woo Suk, Tae Ph.D
Seoul, Korea
|