[Eeglablist] ICA preprocessing and bipolar montage configuration

[Scott Makeig]

Jason -  But the mapping problem remains, no? Will ICA make the IC maps
(including somehow the bipolar channel) smooth even when the data contains
a bipolar channel?  What about this argument:  Making such a dataset
zero-mean does not necessarily mean that the difference between the common
reference and each electrode in the bipolar channel is 0.  If the sources
were [equally everywhere], then this might be the case, but in actual fact,
is it?

Scott

On Wed, Jun 1, 2022 at 12:07 PM Jason Palmer <japalmer29 at gmail.com> wrote:

> A bipolar montage is basically just a linear transformation, and assuming
> you use at most nbchan number of bipolar channels, it is invertible.
>
>         x = A*s
>         B*x = B*A*s = M*s
>         A = inv(B) * M
>
> Where B is the bipolar montage with 1 -1 in the columns corresponding to
> the
> bipolar difference for each row, and M is the EEG.icawinv after running ica
> on the bipolar transformed data.
>
> Jason
>
> -----Original Message-----
> From: eeglablist [mailto:eeglablist-bounces at sccn.ucsd.edu] On Behalf Of
> Scott Makeig
> Sent: Wednesday, June 1, 2022 8:52 AM
> To: Velu Prabhakar Kumaravel <velu.kumaravel at unitn.it>
> Cc: EEGLAB List <eeglablist at sccn.ucsd.edu>
> Subject: Re: [Eeglablist] ICA preprocessing and bipolar montage
> configuration
>
> Velu -
>
> ICA is a linear decomposition, so should be able to decompose bipolar and
> common-reference channels in the same session together. However, the
> bipolar
> channels will be 'floating' with respect to the common-reference channels.
> This might interfere with the decomposition (I have no practical experience
> here) - but even if not it will mean that the ICA scalp maps should not be
> plotted to include the bipolar channels. Here I imagine you might try to
> find fixed offsets representing a 'standing difference'
> between each bipolar channel and the common reference channel that would
> produce max smooth  IC maps -- but are these differences reliably
> stationary? I'd be interested to see a result of attempting this ...
>
> Scott Makeig
>
> On Wed, Jun 1, 2022 at 11:42 AM Velu Prabhakar Kumaravel <
> velu.kumaravel at unitn.it> wrote:
>
> > Dear EEGLABers,
> >
> > Does anyone know the effects of ICA preprocessing on EEG acquired
> > using bipolar configuration?
> > I tried on a few datasets and it looks like the decomposition is not
> > effective. Classifying using ICLabel results in more number of "Other"
> > category.
> >
> > Could someone provide insights on this?
> >
> > Best regards,
> >
> > Velu Prabhakar Kumaravel, Ph.D. Student Center for Mind/Brain
> > Sciences, University of Trento, Italy
> > _______________________________________________
> > 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
> >
>
>
> --
> Scott Makeig, Research Scientist and Director, Swartz Center for
> Computational Neuroscience, Institute for Neural Computation, University of
> California San Diego, La Jolla CA 92093-0559, http://sccn.ucsd.edu/~scott
> _______________________________________________
> 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
>
>

-- 
Scott Makeig, Research Scientist and Director, Swartz Center for
Computational Neuroscience, Institute for Neural Computation, University of
California San Diego, La Jolla CA 92093-0559, http://sccn.ucsd.edu/~scott