[BIC-announce] PET CBF stats analysis using FMRISTAT (!)

[Keith Worsley]

There's a new web page for PET CBF stats analysis using FMRISTAT at
http://www.math.mcgill.ca/keith/fmristat/index_pet.htm
also reachable from the FMRISTAT home page. It does a stats anlysis similar 
to DOT, but with the latest advances in stats theory (DOT has not been 
changed in over 10 years!).

The main tool is multistat, which is used to smooth the sd of the data to 
obtain ~100 df, somewhere between a voxel sd (no smoothing) and pooled sd 
(infinite smoothing, as currently used by DOT). Stat_summary takes care of 
the P-values, as for fMRI data, using the latest advances in random field 
theory. Transforms to Talairach space, and spatial smoothing (if desired, 
though I don't reccomend it) must still be done outside FMRISTAT before 
starting the analysis.

Thanks to Manou Vaffee for inspiring me to do it, and for supplying the 
data!

Keith Worsley
Department of Mathematics and Statistics, McGill University
805 ouest, rue Sherbrooke, Montreal, Quebec, Canada H3A 2K6
tel: (514) 398-3842, fax: (514) 398-3899
e-mail: keith.worsley at mcgill.ca, web: http://www.math.mcgill.ca/keith