ika16 wrote:
> Dear SPMers
>
> I want to plot t-maps using the results of the brain image data.
> Would anyone give a right direction to get t-maps?
>
> What I understand that before getting t-maps I should make t-contrasts first.
> Is it right?
Yes, but you first need to design and estimate a model. Doing so will
create the SPM.mat file required below...
> Then I tried : tasks->stats->contrast manager in SPM graphics window.
> When I clicked “contrast manager?SPM ask to select SPM.mat file.
> What kind of SPM.mat file I should make?
> I got 4 groups and each group has 5 independent components, which has 510340
> voxels on each independent component. I want to compare different groups each
> components.
> In ‘Contrasts Sessions -> T contrast vector? what kind of T-contrasts vector I
> should put in here? Just showing activation and deactivation the areas of the
> SPM t-map, the contrast will be [ 1 ? 0 ?] ? A 1 x n vector should be entered
> for T- contrasts. n is the voxel number is right? If I am right, how can I put
> it in?
n here is just the number of regressors (aka explanatory variables, or
covariates) i.e. the number of columns in your GLM design matrix. So
for example, a simple two-group t-test would have two columns in the
design matrix, allowing a mean for each group, then you'd have a
contrast like [1 -1] to test group1>group2.
For your design I guess you'll have four columns for group, five for
IC, and various 1x9 contrasts to test the effects and interactions
that you're interested in. (possibly you'll have overall mean, and/or
other regressors in there too, so maybe 1x10 or bigger contrasts, but
you won't have as many columns as voxels -- if you did, you'd never be
able to estimate the model!)
> Any recommend papers to understand about SPM t-map and any advice on this
> matter would be greatly appreciated. I really need help…….
Start here:
http://www.fil.ion.ucl.ac.uk/spm/doc/
There is a quickstart guide, and a wiki linked from there which should
get you going.
There is also a manual under spm5/man/manual.pdf (linked from above
website too).
The HBF book has great chapters on the GLM, contrasts, etc:
http://www.fil.ion.ucl.ac.uk/spm/doc/books/hbf2/
Hope that helps,
Ged.
|