Re: [Marsbar-users] ROI extraction: design matrix based of raw data based?

[Matthew B. <mat...@gm...>]

Hi,

> The question is "When I choose SPM design do I have time courses extracted
> after confounds (realignment parameters and others) are regressed out?" Can
> I choose any contrast to control regressors to regress out?

If you use the extract data options, the only processing that is
applied to the raw data is (optionally with the full options) the
image-per-image scaling applied by SPM - so no confounds are removed,
no filtering is applied.

If you want to do that you should use the marsbar spm graph option on
the results menu.  This is just a straight copy of the SPM function,
and if you choose fitted and adjusted data, and a contrast that
includes only the effects you are interested in, then the resulting
plot will leave these variables in the matlab workspace:

%          Y      - fitted   data for the selected voxel
%          y      - adjusted data for the selected voxel
%          beta   - parameter estimates
%          Bcov   - covariance of parameter estimates
%          cbeta  = betas multiplied by contrast

So - I think you want 'y' here - the data adjusted for all effects of
no interest.  The filtering has also been applied.

Best,

Matthew