[Eeglablist] ASR and blink artifacts

[Tarik S Bel-Bahar]

Hello Nike, some notes below, hope you find them useful!


********************************************
Generally try to understand the ASR method better, as it seems to estimate
the space, and "remove" outlier components, and then rebuilds the data.
Users should remember that the tool finds and uses what it thinks are the
"cleanest" periods in order to estimate how to "change" the dirty periods.

a. asr can remove anything that is weird. Check out the documentation and
also the article on PREP pipeline which uses ASR functions too in the same
way.

b. if you run ica after ASR, and still see an IC component that is related
to eye artifacts such as blinks or lateral eye movement, you could choose
to remove those ICs. However, i have not seen publications explaining how
the ASR cleaning impacts IC decompositions later on in an analysis
pipeline.

If you have not had a chance to, review Makoto's online page of
recommendations, which include steps/recommendations related to ASR and
ICA, I believe.

c. for not removing blinks but just big artifacts, you can hack into the
functions that you are running, and feed it all the channels except
channels near the blink.
However, it's not clear how using ASR to reconstruct only some of the
channels works, as some of the channels and periods will be rebuilt,
whereas other channels will apparently remain untouched. You may want to
check with Dr.Kothe, the developer of that code.

d. you may also examine how the cleaning changes with different settings,
and you can also try running some of the sub-functions by themselves
i believe that there is a clean_rawdata but it calls several subfunctions
such as clean_channels, etc...
So, for example, you could just/only run the function that replaces bad
periods and not run the function that cleans channels, or alternatively you
can set the threshold for channel dropping to be very low, so it does not
drop any.
You can see what functions are called by examining the code and
documentation for the main function you are running.
The documentation is quite helpful if you have not had a chance to look
through, and there is also documentation for the subfunctions.











On Wed, Apr 5, 2017 at 1:58 PM, Nike gnanateja <nikegnanateja at gmail.com>
wrote:

> Dear List,
>
> I have been trying to use clean_rawdata (ASR) to clean the EEG. I tried
> using the default settings. On using the vis_artifacts I saw that most of
> the blinks have been reduced drastically following clean_rawdata.
>
> Here are my questions:
>
> 1) Is ASR supposed to remove eyeblinks, or am I doing something wrong ?
>
> 2) If ASR removes the eyeblinks, should I or should I not use ICA to
> remove the blink artifacts ?
>
> 3) Is there a way to make sure that ASR just removes huge artifacts like
> head movements, channel noise or loose contacts etc., without affecting the
> blinks, which can be later removed using ICA ?
>
>
>
> Nike
>
> --
> G Nike Gnanateja <http://goog_636235333>
> Junior Research Fellow,
> Department of Audiology,
> All India Institute of Speech
> and Hearing Mysore-06
>
> _______________________________________________
> 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/20170405/c4e31897/attachment.html>