[Eeglablist] Can I do coherence between independent components?

[julien Lagarde]

Rachel,

directly motivated by the question you raised,the tdsep
algorhythm, a second order blind separation (SOBI),
minimizing second order statistical dependence simultaneously
taking different time delays (SOBI is a function implemented in EEGLAB),
has been tested in the following paper by Meinecke and col 2005:

Meinecke FC, Ziehe A, Kurths J, Muller KR.

Measuring phase synchronization of superimposed signals.

Phys Rev Lett. 2005 Mar 4;94(8):084102

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=
Retrieve&db=PubMed&list_uids=15783894&dopt=Citation


The paper doesn't provide an in-length, theoretical treatment,
as you are asking for; however, the results obtained on simulated
data indicate that indeed two synchronous but distinct sources of
activity can be efficiently separated via a SOBI solution;

It seems that the method fails only if the phase and amplitude
are perfectly synchronized, which is not the case in the brain,
where perturbation by noise, phase shift (synchronization not perfectly
centered at zero phase difference), and also synchronization of phases in
the absence of synchronization of amplitudes, altogether are more the rule
than the exception.

the question remains quite intriguing.

julien lagarde






> Hi, Dear All:
>
> I am recently working on ICA. One question confused me: could I do
> coherence analysis between independent components of EEG?
>
> Because ICA assume sources are mutually uncorrelated. So after ICA, the
> separated sources should be uncorrelated. However, some people used ICA
> and then did coherence analysis among components. What is the basic theory
> for doing that? Could you recommend some papers for their theory?
>
> Thank you very much.
>
> Rachel/
> _______________________________________________
> eeglablist mailing list eeglablist at sccn.ucsd.edu
> Eeglablist page: http://sccn.ucsd.edu/eeglab/eeglabmail.html
> To unsubscribe, send an empty email to
> eeglablist-unsubscribe at sccn.ucsd.edu
>


julien Lagarde PhD

http://julienlagarde.free.fr

@ Movement Science Dept
Faculté des Sciences du Sport                        tel: (33) 4 67 41 57 35
700, Av. du Pic St Loup
34090 Montpellier, France
http://www.edm.univ-montp1.fr/
http://www.enactivenetwork.org
@ Human Brain Behav Lab, Boca Raton fl usa,
Center for complex systems brain sciences
http://www.ccs.fau.edu/section_links/HBBLv2/