> fiddling around with the imaging data directly provided by the ICBM
> (from www.loni.ucla.edu/ICBM/ICBM_Probabilistic.html), I wondered if the
> difference in the scale factors will confuse spm if this data is used as
> "custom priors" (the data contains pixel values from 0 - 32767). I would
> appreciate your feedback.
I have also been playing with this data, and it would confuse SPM. The
scalefactors need to be changed so that the probabilities are in the range
of 0-1. The space of these data are also different (note that when the
origin field is set to zero, and there is no .mat file, SPM assumes that
the origin is at the centre of the volume). They are also smaller than
the MNI datasets, because they reflect the average sizes of the subjects'
brains. The affine registration part of SPM2 has been tweeked in order
to encode the amount of zooming needed to get a typical brain as large as
the MNI data. If you want to affine register with these data, then you
could consider changing the spm_defaults.m file from:
defaults.normalise.estimate.regtype = 'mni';
defaults.segment.estimate.affreg.regtype = 'mni';
to:
defaults.normalise.estimate.regtype = 'subj';
defaults.segment.estimate.affreg.regtype = 'subj';
This biases the zooming towards a factor of 1, rather than the values
needed for resizing to the MNI space.
All the best,
-John
--
Dr John Ashburner.
Functional Imaging Lab., 12 Queen Square, London WC1N 3BG, UK.
tel: +44 (0)20 78337491 or +44 (0)20 78373611 x4381
fax: +44 (0)20 78131420 http://www.fil.ion.ucl.ac.uk/~john
|