[Eeglablist] ICA for removal of EOG artifacts

[arno]

> Since there were 2 different conditions in my recordings 
> (actual-movement and imagined-movement), should I run ICA on the whole 
> 5 minutes continuous data (with both conditions) or should I run it 
> for concatenated epoched data (2.5 minutes for each condition) even 
> though the data points were abit insufficient?
>
You should definitely run ICA on the continuous data with both condition 
(or data epochs concatenated from the two conditions). If you do not do 
that and run ICA on each condition separately, reviewers of your paper 
may always argue that you have a difference between the two conditions 
because you removed different ICA components (they will probably be very 
similar but probably not perfectly identical) in the two conditions.

For removing artifacts, the exact number of points is not so critical. 
It is much more critical if you are interested in obtaining ICA 
components representing EEG source activity. Also the number of data 
points given in the tutorial is approximative and based on personal 
experience with real data. It is not derived from a mathematical formula.

Here are some related threads.

http://sccn.ucsd.edu/pipermail/eeglablist/2006/001570.html

http://sccn.ucsd.edu/pipermail/eeglablist/2006/001485.html

Hope this helps,

Arno