...although, if I may clarify further, even if you care only about interactions between two very specific regions/sources, it's generally advisable to use a multivariate approach which incorporates the largest number of sources possible. This will help reduce the risk of false positives incurred by purely bivariate methods (such as coherence) due to correlation of the two measured (included) variables with unmeasured (excluded) variables. Partial coherence is one such multivariate estimator for correlation (in the frequency domain).<div>
<br></div><div>Tim <br><br><div class="gmail_quote">On Mon, Oct 15, 2012 at 10:11 PM, Tim Mullen <span dir="ltr"><<a href="mailto:mullen.tim@gmail.com" target="_blank">mullen.tim@gmail.com</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
well, our (not yet released) Hierarchical Bayesian Modeling approach is geared specifically towards multi-subject (second-level) analysis with possible missing data. Iman's question is what to do if, in a <i>single</i> subject, there is more than one source in a given region of interest. How shall he select "the appropriate source?" My answer is another question: what does one mean by "the appropriate source?" For instance, there are many possible sources of EEG activity in the frontal cortex. If a subject produces multiple frontal components, the decision as to which to select depends on the specificity of the hypothesis being tested. If your hypothesis is restricted to a specific anatomical region in the frontal cortex, then you might localize the components -- using your choice of localization method (e.g. equiv. current dipole, minimum-norm, or beamforming) -- and select any component which localizes to this region. If your hypothesis is more general (e.g. you care only whether any frontal region interacts with any parietal region) then you should enter <b>all</b> candidate frontal and parietal sources into a <b>multivariate</b> model and examine the <b>partial</b> coherence (or perhaps multivariate granger causal, if you care about directed relationships) relationships between your variables of interest. As Makoto mentioned, this sort of thing is entirely possible in the SIFT framework. The handbook has some more details and references.<div>
<br></div><div>Tim<br><div><br></div><div><div><div><br></div><div><div><div><br><div class="gmail_quote">On Mon, Oct 15, 2012 at 9:36 PM, Makoto Miyakoshi <span dir="ltr"><<a href="mailto:mmiyakoshi@ucsd.edu" target="_blank">mmiyakoshi@ucsd.edu</a>></span> wrote:<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">Dear Iman and ICA persons,<br>
<div><br>
> So, how can I pick up topo and source time series<br>
> among all 128 sources<br>
<br>
</div>Jason and Scott, this Iman's question is legitimate. I have been<br>
asking the same question too. Other than Tim's new implementation of<br>
the Bayesian stuff in SIFT in future, what other conventional approach<br>
can help?<br>
<div><div><br>
Makoto<br>
<br>
<br>
2012/10/15 Iman M.Rezazadeh <<a href="mailto:irezazadeh@ucdavis.edu" target="_blank">irezazadeh@ucdavis.edu</a>>:<br>
> Dear Makoto and Tim,<br>
> Thanks for your reply and suggestion ! I will definitely go through SIFT to<br>
> see how can it help me ...<br>
> Each source (after BSS) has a time series and a topo map. Suppose that I<br>
> have 128 channels means 128 sources and I want to know which sources are<br>
> related to each other. So, how can I pick up topo and source time series<br>
> among all 128 sources ( after removing some sources in the artifact<br>
> rejection stage ) ?<br>
> - for an example, I am interested to see if there is any fronto-parital<br>
> synchrony for an inhibitory task. After BSS, there are more than 1 source<br>
> which have frontal or parietal major activities in their topo. Now how could<br>
> I select the appropriate source?<br>
><br>
> I have found the attached paper. Do you have any thought?<br>
> Best,<br>
> Iman<br>
><br>
> Iman M.Rezazadeh, PhD<br>
> Postdoctoral Fellow<br>
> Center for Mind and Brain<br>
> University of California, Davis<br>
> <a href="mailto:irezazadeh@ucdavis.edu" target="_blank">irezazadeh@ucdavis.edu</a><br>
> cell:<a href="tel:310-490-1808" value="+13104901808" target="_blank">310-490-1808</a><br>
><br>
><br>
><br>
> -----Original Message-----<br>
> From: Makoto Miyakoshi [mailto:<a href="mailto:mmiyakoshi@ucsd.edu" target="_blank">mmiyakoshi@ucsd.edu</a>]<br>
> Sent: Monday, October 15, 2012 7:53 PM<br>
> To: Iman M.Rezazadeh<br>
> Cc: <a href="mailto:eeglablist@sccn.ucsd.edu" target="_blank">eeglablist@sccn.ucsd.edu</a><br>
> Subject: Re: [Eeglablist] Coherence or Correlation among set of electrodes<br>
><br>
> Dear Iman,<br>
><br>
> I wonder why Tim did not mention his SIFT.<br>
> <a href="http://sccn.ucsd.edu/wiki/SIFT" target="_blank">http://sccn.ucsd.edu/wiki/SIFT</a><br>
> Apply SIFT after ICA. You can choose ICs so that they cover areas of<br>
> interests.<br>
><br>
> Makoto<br>
><br>
> 2012/10/15 Iman M.Rezazadeh <<a href="mailto:irezazadeh@ucdavis.edu" target="_blank">irezazadeh@ucdavis.edu</a>>:<br>
>> Hi,<br>
>><br>
>> Just wonder if there is a way to calculate the coherence measure<br>
>> between two regions( set of channels-instead of two single channels)<br>
>> in EEGLAB or any other software? In other word, how can we find<br>
>> regions in the brain which their activities are mostley related to each<br>
> other using EEG ?<br>
>><br>
>><br>
>><br>
>> Best,<br>
>><br>
>> Iman<br>
>><br>
>><br>
>><br>
>> Iman M.Rezazadeh, PhD<br>
>><br>
>> Center for Mind and Brain<br>
>><br>
>> University of California, Davis<br>
>><br>
>> <a href="mailto:irezazadeh@ucdavis.edu" target="_blank">irezazadeh@ucdavis.edu</a><br>
>><br>
>> cell:<a href="tel:310-490-1808" value="+13104901808" target="_blank">310-490-1808</a><br>
>><br>
>><br>
>><br>
>><br>
>> _______________________________________________<br>
>> Eeglablist page: <a href="http://sccn.ucsd.edu/eeglab/eeglabmail.html" target="_blank">http://sccn.ucsd.edu/eeglab/eeglabmail.html</a><br>
>> To unsubscribe, send an empty email to<br>
>> <a href="mailto:eeglablist-unsubscribe@sccn.ucsd.edu" target="_blank">eeglablist-unsubscribe@sccn.ucsd.edu</a><br>
>> For digest mode, send an email with the subject "set digest mime" to<br>
>> <a href="mailto:eeglablist-request@sccn.ucsd.edu" target="_blank">eeglablist-request@sccn.ucsd.edu</a><br>
><br>
><br>
><br>
> --<br>
> Makoto Miyakoshi<br>
> JSPS Postdoctral Fellow for Research Abroad Swartz Center for Computational<br>
> Neuroscience Institute for Neural Computation, University of California San<br>
> Diego<br>
<br>
<br>
<br>
--<br>
Makoto Miyakoshi<br>
JSPS Postdoctral Fellow for Research Abroad<br>
Swartz Center for Computational Neuroscience<br>
Institute for Neural Computation, University of California San Diego<br>
</div></div></blockquote></div><br><br clear="all"><div><br></div></div></div><span><font color="#888888">-- <br>--------- αντίληψη -----------<br>
</font></span></div></div></div></div>
</blockquote></div><br><br clear="all"><div><br></div>-- <br>--------- αντίληψη -----------<br>
</div>