[Eeglablist] ICA on continuous v. epoched data?

[Grega Repovs]

Andrew,

Once you have computed ICA, you have two views od your data available. One is the view of the original data channels that you can plot and analyze in different ways, the other is of the estimated components, virtual channels that represent the estimated sources, that you can also view and analyze. In EEGLAB you can look at the component channels and ERPs through "Plot" menu (e.g. "Component activations (scroll), Component maps, Component ERPs …) and do all that you can otherwise do with regular channels.

To "compare the involvement of the identified components between conditions", you can, for instance, compute component ERPs for different conditions in your study (represented by different sets of epochs) and compare them, the same way you would do it for regular channels.

Best,

Grega


On Mar 31, 2011, at 6:01 AM, Andrew Hill wrote:

> Oh - and I didn't mean to ignore Grega's reply - didn't realize it was there until now!  Thanks for clearing up some of my confusion, Grega.
> 
> That being said - conceptually it makes complete sense, but not at all sure how to do this part:
> 
>>  Given that, the best strategy seems to be to go for one ICA solution for the whole recording and then compare the involvement of the identified components between conditions.
> 
> I think I need a "functional" as opposed to theoretical pointer.
> 
> Best,
> Andrew
> 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://sccn.ucsd.edu/pipermail/eeglablist/attachments/20110331/9b6adb80/attachment.html>