[Eeglablist] uninterpretable component

[Baris Demiral]

Yes, I agree with Stefan,

Also I have some other comments: You better not include particularly M1, M2,
inferior-ocular (below the eye) VEOG in the ICA. This is mainly because of
the fact that they are not really very relevant/informative/strong for the
cognitive based ICAs. Inf-Ocular is very noisy most of the cases and M1 and
M2 have very small potential and small representative power.

So such electrodes are "distortions" for the ICA: they make ICA baffle. Do
you still really want to include the other EOG electrodes? Then maybe just
include Supraocular VEOG (above the eye) and maybe the HEOGs (if possible
include both right and left sides because they are not showing symmetrical
activities: When you move your eye to the right, right HEOG gets peaky
positive and the left HEOG becomes slow-drifty negative).

But even for the detection of the artifact related ICAs you still can do
without the VEOG and HEOG just by looking at the ICA components over the
edgy AF8, AF10 electrodes etc. I suppose many ICA based correction
algorithms do not need these electrodes. But some programs (like Analyzer2)
also asks VEOG and HEOG though, maybe for better use of these activities in
a particular algorithm. I do not personally include these electrodes in my
ICA based corrections or ICA decompositions. I only use them maybe for
classical artifact rejection over epochs if I need to.

Baris

2011/3/24 Stefan Debener <stefan.debener at uni-oldenburg.de>

>  Dear Iren,
>
> It is difficult to evaluate the quality of a single IC without further
> information. How do the other ICs look like? Do you find many dipolar ICs,
> and reasonable artefact ICs? If not I would suspect that the overall quality
> of the decomposition is rather poor. Consider HP filtering and pruning your
> data first (see EEGLAB tutorial).
> Also, are you sure you have the correct electrodes location file? It seems
> you have recorded against the left mastoid. The plot suggest one left
> lateral HEOG and one right inferior VEOG channel. We prefer not using
> bioloar recordings anymore, recording all your data against a common
> reference is usually easier (to collect and later process) and reveals
> basically the same amount of information. The evaluation of topographies is
> also complicated by a mixture of unipolar and bipolar channels, so this
> could explain some prominent features of the map you are showing.
>
> Best,
> Stefan
>
>
> Am 3/22/11 3:03 PM, schrieb Barkaszi Irén:
>
> Dear all,
>
> I got an uninterpretable component in my analysis. This component's scalp
> map is very  diffuse and the source dipole is outside the brain. I would
> like to quantify the component's quality with a single equivalent current
> dipole but it would entail the elimination of the component, though its ERP
> contributes strongly to the data ERP. Do you think it is correct to throw it
> out?
> Please find enclosed a figure about this component.
>
> Thank you in advance,
> Iren
>
>
> _______________________________________________
> 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
>
>
>
> --
> Prof. Dr. Stefan Debener
> Neuropsychology	Lab
> Department of Psychology
> University of Oldenburg
> D-26111 Oldenburg
> Germany
>
> Office: A7 0-038
> Phone: +49-441-798-4271
> Fax:   +49-441-798-5522
> Email: stefan.debener at uni-oldenburg.de
>
>
> _______________________________________________
> 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
>



-- 
SB Demiral, PhD.
Department of Psychology
7 George Square
The University of Edinburgh
Edinburgh, EH8 9JZ
UK
Phone: +44 (0131) 6503063
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://sccn.ucsd.edu/pipermail/eeglablist/attachments/20110325/5f7340ee/attachment.html>