[Eeglablist] Electrode placement and ICA

[Scott Makeig]

Tom,

What counts is the distribution of independent sources projecting to the 
electrodes. Examining the ICA component maps (Plot > Component Maps > In 
2-D) and equivalent dipole locations (Tools > Locate dipoles using 
DIPFIT > Autofit components) shows that sources comparatively far from 
the posterior electrodes also project to them. To separate out these 
contributions, it seems best to give ICA a 'good view' of all the 
components that contribute -- as well as of all the components that 
contribute to those components (e.g., eye movements). [This surmise 
might be formally tested first by simulation and then by experiment with 
a head phantom].

This observation also suggests that in general that more channels is 
better (we are certainly pleased with our 32-channel results, published 
in the Science paper in 2002, but are even more pleased with results 
from our current 72-channel data -- and are awaiting hard evidence from 
our new 256-channel system...).

Scott

busey at indiana.edu wrote:
> I'm setting up a new 32 channel system and I need to decide on electrode 
> location placements. I do mainly visual components at around 100-300 ms that 
> tend to be distributed near the back of the head. Does the ICA procedure 
> benefit from equally-spaced electrodes over the entire head, or is it better to 
> cluster them in the region of interest? If I do this, I will have unequal 
> spacing, since I will have some frontal electrodes. Is there something magic 
> about equidistant electrode placement for ICA?
> 
> 
> Thanks,
> 
> Tom Busey
> 
> busey at indiana.edu
> _______________________________________________
> Eeglablist mailing list Eeglablist at sccn.ucsd.edu
> Eeglablist page: http://www.sccn.ucsd.edu/eeglab/eeglabmail.html
> To unsubscribe, send an empty email to eeglablist-unsub at sccn.ucsd.edu
> To switch to digest mode, send an empty email to eeglablist-digest at sccn.ucsd.edu