[Eeglablist] Using ICA for labelling artifacts

[Velu Prabhakar Kumaravel]

Thanks a lot for your comments on this topic.

I reduced the number of channels to 4 (from the original 21) to adapt ICA
for short data (as suggested by Onton et al., 2006). Accordingly, for a
4-channel EEG, we would need 20*4*4 = 320 sample points whereas I had used
1024. There is a limitation on the number of channels, too, which is
perhaps not yet discussed.

Thanks, Scott, for pointing out that ICLabel is trained for (high-density)
scalp maps and for proposing the alternate solution using traditional ICA
removal tools (Post-AMICA is interesting!)

Best,

Velu



On Tue, 19 Jul 2022 at 21:29, Cedric Cannard <ccannard at protonmail.com>
wrote:

> Hello,
>
> My 2 cents are:
> 1) ICA does not perform well with few channels (I don't remember the
> minimum) and lowpass-filtered data (keeping high-frequency helps the
> algorithm identify high-frequency bursts).
>
> 2) I would include other artifactual components if the goal is to
> discriminate between "Clean" and "Artifactual".
>
> > From our preliminary analysis, more than 50% of windows are labeled for
> Artifact, hence the concern.
>
> Are you saying you are running ICA on each 4 s epoch?? No wonder it
> doesn't work. The algorithm needs data to learn and separate components.
>
> Cedric Cannard
>
>
>
> ------- Original Message -------
> On Tuesday, July 19th, 2022 at 5:45 AM, Velu Prabhakar Kumaravel via
> eeglablist <eeglablist at sccn.ucsd.edu> wrote:
>
>
> > Hello,
> >
> > Does anyone have some thoughts to share on this?
> >
> > Thanks,
> >
> > Velu
> >
> > On Thu, 14 Jul 2022 at 12:00, Velu Prabhakar Kumaravel <
> > velu.kumaravel at unitn.it> wrote:
> >
> > > Dear EEGLABers,
> > >
> > > As we know, it is rare to find data labeled for artifacts which makes
> it
> > > hard to develop or validate artifacts detection algorithms. My
> colleagues
> > > and I wonder if we could use ICA followed by ICLabel to label the
> segments
> > > of data as "Artifact" or "Clean". The steps are as follows:
> > >
> > > 0) To facilitate reliable ICA decomposition for short windows of data,
> > > only 4 channels are kept (known apriori)
> > > 1) Bandpass filter (1-45 Hz)
> > > 2) Segment data into 4s non-overlapping windows (sampling rate = 256
> Hz,
> > > 1024 samples)
> > > 3) Perform ICA
> > > 4) Classify components using ICLabel
> > > 5) If there exists at least 1 eye or muscle Component with probability
> >
> > > 0.95, the given window of data is "Artifact". Otherwise, "Clean".
> > >
> > > Do we see any potential problems in the approach? From our preliminary
> > > analysis, more than 50% of windows are labeled for Artifact, hence the
> > > concern.
> > >
> > > I appreciate your feedback.
> > >
> > > Thanks and 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
>