Dear SPM'ers,
I am painfully aware my query may have been dealt with in early incarnations
of SPM but I hope someone, somewhere will take pity.
I've been using SPM99 (PET models) to estimate changes in gradient echo (GE)
images of small animals under the assumption that GE acquired volumes are
independent measures (with negligible steady state effects between volumes)
and so akin to PET data. I believe (hope) the only issue with using such
models comes at the point of correcting for multiple comparisons. The area
of our work is pharmacological MRI so we have little a prior knowledge of
'expected' changes and so are forced to correct for the entire volume.
My questions then are:
What are the point spread functions for transforming into resel space for
spm99 PET models? I've come across a figure of 20mm (Worsely K.J. et al
(1996) A Unified Statistical Approach for Determining Significant Signals in
Images of Cerebral Activation. HBM 4:58-73.)
Are the values easily changed in the SPM code?
Is this all rubbish, my assumptions being completely incorrect?
Any help would be most gratefully received,
Andrew Lowe
Research Assistant/PhD Student
Neuroimaging Research Group - Neurology
Institute of Psychiatry
P042
ASB
4 Windsor Walk
London
SE5 8BB
020 7848 0424
020 7848 0055 (fax)
[log in to unmask]
www.iop.kcl.ac.uk
|