[Eeglablist] Two step source connectivity analysis (as implemented in SIFT)

James Jones-Rounds jj324 at cornell.edu
Fri Feb 7 14:12:15 PST 2014


Hi Imali,

This is a fascinating question, but hopefully one that's easily resolved.
My understanding of ICA is that while it does try to find maximally
independent sources in the channel-space matrix, it is primarily a spatial
filtering method. Therefore, you could have dependent or interdependent
brain regions (e.g. Region A and Region B) show up as different independent
components, because they are spatially localized to different areas of the
brain. Furthermore, unless they are truly simultaneously active, then there
will likely be at least some temporal delay between their activity patterns
that will further help them dissociate into unique ICs. I think in these
two ways, it is possible to measure causal information flow, connectivity,
and coherence among independent sources calculated using ICA.

I'm curious to see what some of the ICA experts with their ear to this list
would say, though!

James

-- 
James Jones-Rounds
Laboratory Manager
Human Development EEG and Psychophysiology (HEP) Laboratory,
Department of Human Development,
--------------------------------------------
Cornell University | Ithaca, NY
607-255-9883
eeg at cornell.edu
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://sccn.ucsd.edu/pipermail/eeglablist/attachments/20140207/3176f12a/attachment.html 


More information about the eeglablist mailing list