[Eeglablist] ICA filtering and IC semi automatic classification

[Maximilien Chaumon]

Hi Alex,

None of the procedures you suggest seems really wrong to me. I would
probably run the ICA on 1Hz data to get a neat decomposition (in my own
experience too, the decomposition works best on 1Hz high pass filtered
data), apply the resulting IC weights on the 0.1Hz data (useful if you're
studying slow potentials), then pass this through SASICA to decide which
components to reject. I've never tried it myself, so I'll be interested if
anyone has comments about that.

As a side note, I do not recommend to blindly reject everything that SASICA
will suggest if you activate all measures. I recommend parsimonious
rejection according to the needs of your experiment (see the paper attached
to the plugin for full explanation).

Best,
Max

Le ven. 13 avr. 2018 à 00:58, Alexandre Obert <obert.alexandre at gmail.com> a
écrit :

> Hi all,
>
> I read somewhere
> <https://github.com/CSC-UW/csc-eeg-tools/wiki/Filtering-and-ICA>that one
> could perform ICA onto data filtered with 1Hz (high-pass) and then apply
> ICA matrix onto 0.1Hz filtered data.
> However, I wonder how to deal with ICs classification algorithms such as
> SASICA or ADJUST in such process?
>
> SASICA onto 1Hz filtered data send different results from SASICA onto
> 0.1Hz filtered one.
> So, what's the best process:
> 1) performing SASICA onto 1Hz data and then reject artifactual ICs in
> 0.1Hz (don't feel comfortable with this)
> 2) forgetting such process and just using ICA and SASICA onto 0.1Hz data
> directly
> 3) following the mentioned process and performing SASICA onto 0.1Hz data
> with ICA matrix from 1Hz data
> 4) something else
>
> Alex
> _______________________________________________
> 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
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://sccn.ucsd.edu/pipermail/eeglablist/attachments/20180413/901c6d0f/attachment.html>