[Eeglablist] Component statistics

[Makoto Miyakoshi]

Dear Paolo,

I don't recommend kurtosis for that purpose.
Open GUI and click 'reject components by map' to show a table of
topos. Pick up ones that is likely to be EOG, EMG, or non-dipolar
ones. Check out their spectra and ERP images by clicking their number.

I recommend you read 'Independent compoenents are dipolar' Delorme et
al., 2012 Plos One.

Makoto

2012/8/29 paolo federico <paolofed at univaq.it>:
> Hello everybody,
>
> I have some problem with the post-ICA analysis. I would detect all artifact
> components and then remove them from my dataset (I am working with an
> epoched dataset).
> I am actually using the tool "Dat statistics --> Component statistics". I
> have a question: Which is the threshold of kurtosis to keep in mind in order
> to have an objective method to discriminate artifact components? Basing on
> my knowledges, when kurtosis = 0 it means that the distribution is normal,
> even if there is a tolerance window (-1/+1). Is it possible that all
> components, even those quite-normal distribuited, have kuortosis values
> higher than 3?
>
> In addition, using Tools-->Rejecting data using ICA-->Reject data (all
> methods)--->Find abnormal values why as default value there are +25/-25  µV
> for, respectively, upper and lower limits? What does it means.
>
> I'm sorry for all these question, but reading the EEGLAB wiki I did not
> understand very much.
>
> Best regards,
>
> Paolo
>
>
> --
> Paolo Federico, PhD Candidate in Biotechnology
> Department of Experimental Medicine, University of L'Aquila
> Via Vetoio, Coppito II, 67100 L'Aquila, Italy
> email: paolofed at univaq.it
>
>
> _______________________________________________
> Eeglablist page: http://sccn.ucsd.edu/eeglab/eeglabmail.html
> To unsubscribe, send an empty email to eeglablist-unsubscribe at sccn.ucsd.edu
> For digest mode, send an email with the subject "set digest mime" to
> eeglablist-request at sccn.ucsd.edu



-- 
Makoto Miyakoshi
JSPS Postdoctral Fellow for Research Abroad
Swartz Center for Computational Neuroscience
Institute for Neural Computation, University of California San Diego