On 8/10/06, Alle Meije Wink <[log in to unmask]> wrote:
> IF the MNI people have use SPM/other nonlinear deformation fields. I
> don't know if this is the case... I thought that the MNI images were
> constructed by rigid/affine registration and averaging.
All the mni models released at
packages.bic.mni.mcgill.ca/tgz/mni-models-* were created using affine
registration only. (9 parameters).
Quick summary:
250 images were aligned using manual identification of points (Evans)
-> initial 250 model created.
55 extra images were added by linear automatic (intensity)
registration. (Colins)
-> average305 model was created and released with mni_autoreg
Colin was scanned a bunch of times for a model with high SNR and definition.
-> colin27 model created.
but average305 had poor coverage at top and bottom of brain
ICBM project started.
152 images from the ICBM project were linearly registered (Colins) to the
average 305 model and averaged using ICBM co-ordinates.
-> ICBM152 model created.
--
Andrew Janke ([log in to unmask] || http://a.janke.googlepages.com/)
Canada->Montreal Cell: +1 (514) 924 2012
|