Dear Sun,
On Mon, Apr 19, 2010 at 8:07 AM, SUBSCRIBE SPM Sun Delin
<[log in to unmask]> wrote:
> Dear Vladimir,
>
> Thanks for your so quick reply! I commented out of two lines (i.e. 'spm_eeg_invert_display(D{i});' and 'drawnow') in spm_eeg_invert.m as you suggested and it does work. I would like to know how to do the following analysis. Should I set time window and get images from each subject and then do the 2nd level analysis?
Yes
> Is the processing of smoothing still needed?
Yes. The unsmoothed images have values just on a 2D manifold in a 3D
volume and this violates the assumptions of random field theory. You
can use minimal smoothing of 8mm, but you must smooth.
> Furthermore, is there a way to review the results of group inversion as reviewing group average results in the sensor space, i.e. just one data file containing 4 conditions, not reviewing subjects one-by-one?
No because there is no such thing. The algorithm constraints the
individual inversions to be similar across the group but it does not
produce a group solution.
Best,
Vladimir
>
>
|