[Eeglablist] Artifact Removal by epoching

[Tarik S Bel-Bahar]

Hello Adeel, notes below, best wishes.

*********************************************
If you would like more specific suggestions besides the below,  send more
details/pictures about the confusion you're experiencing.

It's likely best develop one's one standardization based on exploring how
different settings change results, copying high-quality published methods,
and following recommendations from experts.
One place to look is in the cleaning settings/choices of default settings
several open "automated" EEG preprocessing toolboxes t (e.g., TAPEEG,
HAPPE, PREP, etc..).
One can also find articles that have data similar to yours, and contact
those researchers for their "settings".
Overall, there do not seem to be established published standards (usually
lab-specific and paradigm specific), and authors are usually not required
to provide specific details about rejection thresholds.

You most likely need to modify the settings (play with them) till you get
thresholds that work well for your data. Sometimes these need to be adapted
further for specific datasets or participants.
For before ICA, the reject continuous data method in the eeglab gui can
work quite well. Within the ICA/EEG  communication cleaning after ICA
happens primarily by removal of ICs.
For single epoch rejection, one can first remove ICs, and then run artifact
rejection tools on the cleaned EEG data epochs, using the IC kurtosis etc
tools.
One should also have  expert review of the data across these steps to make
sure that  the data is being cleaned correctly.

For more info, try the following, if you haven't had a chance to yet:
Chapter 01 Rejecting Artifacts of the eeglab online tutorial i
full EEGLAB tutorial and tutorial data  & working with other data (trying
to get it clean)
Googling eeglablist for your topic
Reviewing articles that automate cleaning of data with and without ICA
(search on Google Scholar)
Reviewing artifact/cleaning chapters in recent EEG handbooks
There are also other artifact detecting tools (in other toolboxes) that you
can explore (e.g., Fieldtrip)
lecture slides from ERPLAB on artifact detection:
https://erpinfo.org/lecture-slides

*********************************************







On Fri, Mar 2, 2018 at 6:41 AM Adeel Khan Shigri <shigriadeel at gmail.com>
wrote:

> Respected Sir ,
> I am a little bit confused in rejecting the artifact by mean of Kurtosis ,
> spectra and probablity in both epoch and ica method .Is there any
> standarization so that we can reject artifact by these methods ?I have
> tried to find the solution from the official site of eeglab but i was
> unable to figure out the problem .Kindly give me a valuable suggestion .
> Truly ,
> Adeel
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://sccn.ucsd.edu/pipermail/eeglablist/attachments/20180829/616e13d5/attachment.html>