[Eeglablist] Dealing with a noisy TP10 channel for re-referencing (average TP9-TP10)

[Boki Wang]

Thank you for your tips Cedric! I really appreciate it! - Boki

On Thu, Feb 11, 2021 at 4:17 PM Cedric Cannard <ccannard at protonmail.com>
wrote:

> Hi Boki,
>
> Either remove that subject from the analysis if you really want to do
> mastoids referencing for the group. Or I would recommend removing your
> artifactual TP10 channel for that subject, interpolate from neighboring
> channels, and re-reference with REST/infinity as it is better than AR with
> less channels (see
> https://urldefense.com/v3/__https://link.springer.com/article/10.1007/s10548-019-00707-x__;!!Mih3wA!Rmf6FHWNmbVk0a85C1sQXIoCdjmeOGt09SegVsX44diU94NMwMQl7Dp2aYS0YBCpYZFVXw$ ).
> You can download the REST plugin through the EEGLAB GUI. I don't think the
> command line is available for REST so you have to do it through the GUI in
> the Tools tab once installed.
>
> And you'd have to use the same referencing method for all your
> participants of course if you choose this method.
>
> Cedric
>
> ‐‐‐‐‐‐‐ Original Message ‐‐‐‐‐‐‐
> On Wednesday, February 10, 2021 5:25 PM, Boki Wang <pwang91 at asu.edu>
> wrote:
>
> > Hi list,
> >
> > I am new to (pre)-processing of resting-state EEG data. I am using a 32
> > channel system, where the reference electrode is the right earlobe.
> During
> > offline analysis, I would re-reference the data to average mastoids
> > (average of TP9 and TP10).
> >
> > But the TP10 channel for one of the participants is contaminated by
> noise,
> > so much so that the power spectrum doesn't look like 1/f function, but a
> > NIKE logo where power for frequencies of 12Hz and beyond starts
> increasing
> > (hope this makes sense). I don't think it's appropriate to use the data
> > from this channel for re-referencing.
> >
> > In this case, what's the best thing to do for re-referencing? Can I
> > interpolate TP10 electrode and still use the interpolated data for
> > re-referencing? Or, if I remove this electrode and use average reference
> > (which isn't great for 32 channel data either right?), can I process this
> > participants' data with everyone else's?
> >
> > Would appreciate any pointers!
> >
> > Thank you!
> >
> > Boki
> >
> > 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
>
>
>