[Eeglablist] Extra components in clean compared to uncleaned EEGs-reliable or not?

Clement Lee cll008 at eng.ucsd.edu
Thu Sep 19 12:52:44 PDT 2019

Hi Tony,

It is more likely that cleaning the data uncovered real brain sources,
instead of injecting artifacts that ICLabel considers as brain sources. You
can check the validity of these components and whether or not they are
artifactual by examining the parameters displayed in IC Label (use this
guideline <https://labeling.ucsd.edu/tutorial/labels>).
It's tempting to assume the same subject would have the same number of
components when performing the same task, but alas, intrasubject
variability exists. If you're asking specifically for this case, yes,
pre-processing parameters can and often will change your results.

Clement Lee
Applications Programmer
Swartz Center for Computational Neuroscience
Institute for Neural Computation, UC San Diego

On Wed, Sep 18, 2019 at 11:37 PM Tzeng, Tony H. <ttzeng at lsuhsc.edu> wrote:

> Hi all,
> I ran the ICA on unclean vs cleaned data for one subject at rest, and I
> found two extra components in the cleaned data (which appear to be neural
> sources based on characteristics, scalp map/ERPimage, etc, and they were
> also identified as "Brain" with the new ICLabel plugin). What would be the
> best explanation for them to not have shown up at all following ICA of the
> raw data? Would it be that they were, somehow, masked in the several noisy
> epochs that were rejected in the cleaned version? Or would it be that these
> new components are actually artifactual in nature as well?
> Are there any suggestions for how to confirm/test the reliability of these
> "new" components? Is this occurrence usual when comparing clean to
> uncleaned EEG data? I was expecting both to have similar number of
> components since they are from the same subject (and almost identical, +/-
> some noisy epochs). Thanks for all of your patience.
> Best,
> Tony
> _______________________________________________
> 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

More information about the eeglablist mailing list