Dear SPMers:
Lets assume we have PET study of n=8 subjects and we are able to find
clusters that survive p=0.001 using SPM96.These clusters (5) are in
neurophysiologically correct regions for our paradigm.
What will be the MINIMUM CLUSTER size (please do not tell me volume
corrected p for size should be used, or for height since I cannot get
more than 8 subjects) that could reasonably be reported keeping in
mind initial resolution of PET scan and processing (stereotactic
normalization ) with the inevitable changes in raw data.
Should we report all of them, even if they have one voxel size? I
have difficulty doing that.
Thanks for your help for this "may be philosophical" question
Didine
ps: SPM nonparametric method is likely to give me identical results
and I have to decide on size threshold
--
Badreddine Bencherif, MD
Department of Radiology
Division of Nuclear Medicine
Johns Hopkins University School of Medicine
601 N. Caroline St. / JHOC 4230
Baltimore, MD 21287-0855
Phone : (410) 614-2787
Pager : (410) 283-2050
Fax : (410) 614-1977
email : [log in to unmask]
|