Dear Spm
SPM grand mean scales across what it considers to be brain (above a
certain cutoff of intensity) for the average brain image volume to my
understanding to scale around 100 for a person's average brain across
the time series which is held by the constant beta 0 term (the last
beta calculated) - even though it does not actually scale to around
100 to my calculation but usually the average value on the beta0 term
within brain map is usually 100-200- a bug or my misunderstanding- for
the average brain.
Even if spm grand mean scaled around a value of 100 for the averaged
brain across all voxels, this does not mean that the amygdala at
someone's voxel cannot be averaged around 80 in one person's time
series and averaged around 90 in another person's time series.
Check out a time course for a voxel one person at a voxel for say the
amygdala or its constant term beta0- the last of the betas calculated
- and compare that to another person at the same voxel. The results
will differ.
Thus, percent signal change is a different calculation than what is
normally done with the beta weights and I am wary of the normal beta
weight calculation in group analyses.
Thanks
Jeff Lorberbaum
On Sat, Feb 23, 2008 at 11:23 AM, Stephen J. Fromm <[log in to unmask]> wrote:
> On Thu, 21 Feb 2008 18:25:16 -0800, Jeanette Mumford <[log in to unmask]>
> wrote:
> <snip>
>
> >As for you second question, I haven't used SPM in a while, but I'm fairly
> >sure that like FSL it grand mean scales the data, so that on average all
> >subjects have approximately the same timeseries average (in FSL this is
> >about 10000). This is necessary to make the higher level analysis valid
> for
> >precisely the reason I think you pointed out. So both the %-signal change
> >and the betas are comparable across subjects.
>
> By default, SPM also does grand mean scaling. In SPM, grand mean scaling
> means: for each session for a given subject, scale the data by the mean
> over all voxels over the entire session. ("Session" means a time series
> when the scanner is running continuously.)
>
> <snip>
>
--
Jeffrey Lorberbaum, MD
Assistant Professor of Psychiatry
Penn State- Hershey Med Center
500 University Drive, PO Box 850
Hershey, PA 17033-0850
Phone 717-531-7655
Pager (direct) 717-531-4311 then 3460
Page operator: 717-531-8521
Fax 717-531-6491
e-mail: [log in to unmask]
|