On 7 Apr 2010, at 18:07, Jatin Vaidya wrote:
Hi,
I have a PET dataset that I have normalized into Talairach space.
Hm - within the SPM framework, you're much more likely to have normalized into MNI/ICBM152 space (see Help on Templates).
When I smooth the data using the default 8mm FWHM, the smoothness of the image appears to increase tremendously. I am using an AFNI program to estimate the smoothness (is there a way to do this in SPM?). When I use the AFNI program, I go from about 10mm FWHM (pre-smoothing) to about 20-25mm FWHM (after smoothing).
This looks like a lot of final smoothness, but maybe not impossibly much. The essential smoothness estimate appears in the SPM results stage in the Graphics window at the bottom right - that's the smoothness of your SPM, and should ideally be somewhat matched to the spatial extent of abnormalities expected. What do you get there?
Isn’t this level of smoothness too high? Should I smooth at a much lower level or possibly leave the data as is (i.e., do no smoothing)?
8mm does not strike me as too much at all, rather the opposite - whether this is adequate will depend on a lot of things (isotope, camera resolution, reconstructed voxel size) but my general gut feeling would be that that's unlikely to be too much smoothing - the question aside why AFNI gives you such high readings (I've never worked with AFNI and couldn't comment).
On the other hand, no smoothing will certainly be a bad idea for multi-subject data - you just won't have the anatomical and functional overlap after a standard SPM normalisation; and in addition there are statistical reasons for smoothing (to ensure a more Gaussian structure of your data).
Hope this gets you going; if not, feel free to email a couple of snapshots of your pre/post smoothing data off list.
All the best,
A
---------------------------------
Alexander Hammers, MD PhD
NEW Affiliation:
Chair in Functional Neuroimaging
Fondation Neurodis
http://www.fondation-neurodis.org/
NEW Postal Address:
CERMEP – Imagerie du Vivant
Hôpital Neurologique Pierre Wertheimer
59 Boulevard Pinel
69003 Lyon
France
Telephone +33-(0)4-72 68 86 34
Fax +33-(0)4-72 68 86 10
Email [log in to unmask]">[log in to unmask];[log in to unmask]">[log in to unmask]
Web (London lab)
http://www1.imperial.ac.uk/medicine/people/alexander.hammers.html
http://www.csc.mrc.ac.uk/Research/Groups/ECN/Epilepsy/
---------------------------------
Other affiliations:
Visiting Reader; Honorary Consultant Neurologist
Division of Neuroscience and Mental Health, Faculty of Medicine
Imperial College London
Room 243, Cyclotron Building
Hammersmith Hospital, DuCane Road, London W12 0NN
and
Honorary Reader in Neurology; Honorary Consultant Neurologist
Department of Clinical and Experimental Epilepsy
National Hospital for Neurology and Neurosurgery/ Institute of Neurology,
UCL
33 Queen Square, London WC1N 3BG
Thanks,
Jatin