[Eeglablist] Should Average Reference Include Eye Channels?

[Matthew Stief]

Thank you Makoto.  I will certainly try rereferencing before and after
epoching and see what difference it makes.  One issue I have been having as
a novice EEG user is on precisely what metrics to use when evaluating such
alternate strategies.  Of course I can just look at the continuous data and
see if it looks like there is more noise, but this is dissatisfying to me.
I take Tarik's point that concerns with eyeblinks contaminating the average
don't make much sense since obviously there will be eyeblink activity in
frontal electrodes as well.  I am therefore inclined to just include them
in everything except for dipole fitting.  Just looking at the continuous
data prior to epoching with the reference as only the scalp electrodes
versus the scalp plus eye electrodes I see very little difference.

Also, to confirm a point that I believe Arnaud made on the list, is it
methodologically valid if I epoch without removing the baseline, run ICA,
and then remove the baseline after the ICA has been run?  It was my
understanding that it was removing the baseline and not epoching per se
that made rereferencing potentially problematic.

The mastoids were not linked physically.  When I imported the BioSemi file
I simply indicated that the average of the two channels should be used as
the reference.



On Fri, Feb 10, 2012 at 4:31 PM, Makoto Miyakoshi <mmiyakoshi at ucsd.edu>wrote:

> Dear Matthew (cc Alejandro),
>
> > if I rereference to the average of only the 128 scalp
> > electrodes, and then run ICA on all 132 electrodes, that ICA will
> > average reference the data to all 132 electrodes again?
>
> Correct. It works in that way.
>
> > I believe I remember reading that referencing after epoching and
> baseline removal
> > is erroneous, so how is that possible?
>
> I notice there may be a slight possibility to bias the data, but it
> would be too much to say it is erroneous. If you are concerned, just
> try average reference before and after epoching and baseline
> correction to confirm differences. I would be rather concerned by
> 'linked mastoids'. I hope they were not linked physically.
>
> Alejandro, you are a specialist on this topic. I would appreciate if
> you could give us any comments.
>
> Makoto
>
>
>
> 2012/2/9 Matthew Stief <ms2272 at cornell.edu>:
> > Thanks to all the responses so far!
> >
> > I am using a 128 channel BioSemi ActiveTwo system, so the eye channels
> > are not bipolar.  I imported with a reference to the linked mastoids,
> > and would like to rereference to the average.  My worry was the one
> > raised by Simon-Shlomo below, which is whether or not the extra
> > artifactual activity in the eye channels would contaminate the other
> > channels.
> >
> > This possibility, I gather, is balanced against the goal of having as
> > many electrodes covering the head as completely as possible to match
> > the assumption of average referencing as closely as possible, as Robin
> > mentioned.
> >
> > Makoto, what is your opinion on these two issues?  I am planning on
> > running ICA, not for artifact detection but to isolate the P1, and was
> > planning on including the eye channels to better separate out those
> > those artifactual sources.  And just to be sure I understand you, you
> > are saying that if I rereference to the average of only the 128 scalp
> > electrodes, and then run ICA on all 132 electrodes, that ICA will
> > average reference the data to all 132 electrodes again?  I believe I
> > remember reading that referencing after epoching and baseline removal
> > is erroneous, so how is that possible?
> >
> >
> >
> > On Thu, Feb 9, 2012 at 5:15 PM, Simon-Shlomo Poil
> > <poil.simonshlomo at gmail.com> wrote:
> >> Dear Matthew,
> >>
> >> I would recommend you not to include your eye channels in your average
> >> reference.  If you still have eye-blinks in these channels, they will
> >> just contaminate your other channels.
> >>
> >> (If you already removed your eye-blinks using ICA, it's probably ok to
> >> include them in the average reference.)
> >>
> >> -Simon-Shlomo Poil
> >>
> >>
> >> 2012/2/9 Matthew Stief <ms2272 at cornell.edu>:
> >>> Greetings!
> >>>
> >>> I am using a 128 channel BioSemi active system, and am importing my
> >>> files using the average mastoids as the reference.  I would like to
> >>> rereference to the average, but am unsure if I should use my four eye
> >>> electrodes.  I read in the archives that this shouldn't be done for
> >>> eye electrodes that do not have the same reference as the scalp
> >>> electrodes, but given that in this case they do is there any reason
> >>> not to do it?
> >>>
> >>> --
> >>> _________________________________________________________________
> >>> Matthew Stief
> >>> Human Development | Sex & Gender Lab | Cornell University
> >>> http://www.human.cornell.edu/HD/sexgender
> >>>
> >>>
> >>> Heterosexuality isn't normal, it's just common.
> >>> -Dorothy Parker
> >>> _______________________________________________
> >>> 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
> >>
> >>
> >>
> >> --
> >> Simon-Shlomo Poil
> >>
> >> Webpage: http://www.poil.dk/s/ and http://www.nbtwiki.net
> >
> >
> >
> > --
> > _________________________________________________________________
> > Matthew Stief
> > Human Development | Sex & Gender Lab | Cornell University
> > http://www.human.cornell.edu/HD/sexgender
> >
> >
> > Heterosexuality isn't normal, it's just common.
> > -Dorothy Parker
> >
> > _______________________________________________
> > 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
>



-- 
_________________________________________________________________
Matthew Stief
Human Development | Sex & Gender Lab | Cornell University
http://www.human.cornell.edu/HD/sexgender


Heterosexuality isn't normal, it's just common.
-Dorothy Parker
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://sccn.ucsd.edu/pipermail/eeglablist/attachments/20120214/6d50a7a5/attachment.html>