Dear Zhi,
X0 should reflect how your time series were processed. In SPM, it
contains the block effects and nuisance variables of the filtered and
whitened design matrix:
Y.X0 = SPM.xX.xKXs.X(:,[SPM.xX.iB SPM.xX.iG]);
and the DCT components of the high pass filter (from spm_filter.m):
xY.X0 = [xY.X0 SPM.xX.K(xY.Sess).X0];
How did you analyse your surface-based data? Cannot you bring over the
relevant variables from that analysis? At one point, it might be simpler
to use SPM to fit a GLM on your surface data.
Best regards,
Guillaume.
On 03/12/2018 21:13, Zhi Li wrote:
> Dear Guillaume,
>
> Thank you very much for your help. I have looked into the codes in
> spm_regions.m. The confounds and filter confounds are extracted from the
> SPM.mat file which should has already been estimated, is that right? The
> confounds and filter confounds are the same between different ROIs. It
> should been calculated based on the whole brain data. May I know how to
> calculate this X0 matrix. Now the functional image data I have are
> surface based, Could I use spm commands to do this calculation? Or what
> commands should I refer to for the algorithm?
>
> Thanks and best wishes,
>
> Zhi
>
--
Guillaume Flandin, PhD
Wellcome Centre for Human Neuroimaging
UCL Queen Square Institute of Neurology
London WC1N 3BG
|