[Eeglablist] IC connectivity analysis in between-subject domain

[robert coben]

I would also be very interested in hearing a discussion about ways of assessing coherence via eeglab. Specifically, related to multivariate forms of surface coherence and types of source coherence. Thank you,

Rob Coben, PhD


On Mar 27, 2013, at 9:45 PM, Knyzev GG <knyazev at physiol.ru> wrote:

> Dear list members,
>  
> I would really appreciate a help in applying a connectivity analysis to IC time courses in multi-subject  samples. So far SIFT is only applicable to single subject data. So I have three questions to (principally) Tim Mullen and all interested or knowledgeable members:
> 1.       Are there some extensions or developments of SIFT that allow multi-subject analysis but are not yet included in the current eeglab. If the answer is yes, is it possible to obtain them in the near future?
> 2.       Please refer me to publications that applied connectivity measures in a multi-subject design to ICs derived individually in each subject. My google scholar search yielded only publications that used either a single-subject design (e.g., Mullen et al., 2011) or group ICA.
> 3.       If there are no positive answers to the two above questions, could you please explain how such analysis could be implemented in principle. After applying either clustering procedure or some other approach, like measure projection, in a multi-subject sample we have a number of IC clusters. One cluster may include several components from one subject and no one from another. There have been recently a discussion on the least how to perform and interpret statistical tests on such clusters. My understanding of this discussion is that there are no so far a simple answer. But here I have another concern. How to obtain cluster-related time course which would allow to extract cluster-related phase information which is needed for connectivity analysis. Suppose one wants to do a network graph analysis on connectivity measures treating each cluster as a node. Of course one may just average the time courses of all cluster’s ICs, but I wonder what such averaged across several ICs belonging to different subjects time course would represent and how meaningful would be its phase information?
>  
> Thank you for your time.
>  
> Best regards,
> Gennady Knyazev
> _______________________________________________
> Eeglablist page: http://sccn.ucsd.edu/eeglab/eeglabmail.html
> To unsubscribe, send an empty email to eeglablist-unsubscribe at sccn.ucsd.edu
> For digest mode, send an email with the subject "set digest mime" to eeglablist-request at sccn.ucsd.edu

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://sccn.ucsd.edu/pipermail/eeglablist/attachments/20130328/3bcb3b01/attachment.html>