[Eeglablist] cz/cpz reference

[Makoto Miyakoshi]

Dear Brian,

I discussed your case with Clement.
Let me sort out points here.

1. Data matrix in EEGLAB (i.e., EEG.data) usually does not contain
zero-filled channels--which means it only contains non-reference channels.
2. EEGLAB can store reference info, and when it does, it is stored as 'ref'
in EEG.chanlocs.ref just by name.
3. I updated reference-reated facts in this page just a week ago
https://sccn.ucsd.edu/wiki/Makoto's_preprocessing_pipeline#Why_should_we_add_zero-filled_channel_before_average_referencing.3F_.2803.2F04.2F2020_Updated.29
In the past communication you cited from 2016 I found my comments, but
please refer to the wiki page linked above for the most updated info.

In re-referencing, the most important thing you want to confirm is that
after you re-reference to a new electrode, re-reference back to the
original to see if that recovered the same result. If it does not,
something is wrong.

Makoto

On Tue, Mar 10, 2020 at 1:18 PM Kavanaugh, Brian <brian_kavanaugh at brown.edu>
wrote:

> Hi Clement,
>
> Thanks for your help. This is great. Correct, Cz is a series of zeros. here
> is the prior post I found :
> https://sccn.ucsd.edu/pipermail/eeglablist/2016/011153.html
>
> The full rank approach sounds great. But silly question, how does it
> identify the reference? I don't see how to specify, especially as this
> differs between my two samples (reference in montage for one sample and
> reference not in montage in another).
>
> Thanks again Clement,
>
> Brian
>
> On Tue, Mar 10, 2020 at 3:40 PM Clement Lee <cll008 at eng.ucsd.edu> wrote:
>
> > Hi Brian,
> >
> > Would you link the archived posts from 2016?
> >
> > Full rank average referencing seems to make most sense here. See
> >
> https://sccn.ucsd.edu/wiki/Makoto's_preprocessing_pipeline#Re-reference_the_data_to_average_.2807.2F18.2F2019_Updated.29
> > In your Cpz-referenced datasets, if you were to plot the signal
> > acquired from Cpz, it would simply be a channel of zeroes.
> > In your Cz-referenced dataset, Cz appears in the EEGLAB montage. Does Cz
> > contain a series of zeros? It should, because that's what referencing to
> a
> > single channel is -- subtraction.
> >
> > It's counterintuitive that setting Cz as a reference before average
> > referencing should produce different results than average referencing
> right
> > away. Why should referencing be non-reversible? Removing a channel and
> > interpolating afterwards for referencing purposes (and not noise
> rejection
> > purposes) seems wrong too.
> >
> > Best,
> > Clement Lee
> > Applications Programmer
> > Swartz Center for Computational Neuroscience
> > Institute for Neural Computation, UC San Diego
> > 858-822-7535
> >
> >
> > On Tue, Mar 10, 2020 at 11:16 AM Kavanaugh, Brian <
> > brian_kavanaugh at brown.edu> wrote:
> >
> >> Hi everyone,
> >>
> >> I have a question about cz/cpz reference.
> >>
> >> We collected eeg in two separate samples. in one sample, cpz was the
> >> reference (and we re-referenced to average after). in the other sample,
> >> cpz
> >> wasn't working, so cz was switched to the reference. bc cz was
> originally
> >> set as a recording electrode and because it does appear in eeglab
> >> montage, I set cz as reference in eeglab and then re-referenced to
> >> average,
> >> similar to eeglab wiki recommendations.
> >>
> >> is this the correct approach? i saw a thread from 2016 that said to
> remove
> >> cz, re-reference to average, and then interpolate cz, but wasn't sure if
> >> this is still best approach.
> >>
> >> after low pass filter, i used cleanline and clean_rawdata. before
> running
> >> ica, i wanted to compare the two samples. oddly, the samples have the
> >> opposite association between gamma power and test performance (r = -.3
> >> vs. + .3). Does this make any sense to anyone?
> >>
> >> Thanks for your help!
> >>
> >> brian
> >>
> >> --
> >> Brian Kavanaugh, PsyD.
> >> Pediatric Neuropsychologist, E. P. Bradley Hospital/Rhode Island
> Hospital
> >> Assistant Professor (Research), Warren Alpert Medical School of Brown
> >> University
> >> _______________________________________________
> >> 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
> >>
> >
>
> --
> Brian Kavanaugh, PsyD.
> Pediatric Neuropsychologist, E. P. Bradley Hospital/Rhode Island Hospital
> Assistant Professor (Research), Warren Alpert Medical School of Brown
> University
> _______________________________________________
> 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
>