[Eeglablist] evaluation of artifact removal performance

[Makoto Miyakoshi]

Dear CJ,

> Perspective 1 would be the performance of the removal algorithm in terms
of
its ability to remove artifacts and perspective 2 would be the performance
of the removal algorithm in terms of its ability to NOT remove desired EEG
signals.

This is the open-access paper. See Suppmement 5 'Post-hoc Validation of EEG
Preprocessing'
https://urldefense.com/v3/__https://academic.oup.com/cercorcomms/article/1/1/tgaa046/5881803?login=true*supplementary-data__;Iw!!Mih3wA!SiWjbTlNQsqkf7DpBm8-F6u4RQC_9uxscQT-9yR_zYDh8ldkGMmK4yArfr951cX9KhYtew$ 

The design is not super elaborated but as a proof of concept it may make a
good example to start with.

Makoto

On Sat, Jan 16, 2021 at 8:30 PM Colin Fera <cfera2922 at sdsu.edu> wrote:

> Greetings,
>
> Given a dataset which includes both EMG and EOG artifacts and
> reference channels I would like to make a comparison of artifact removal
> algorithms from two perspectives.
>
> Perspective 1 would be the performance of the removal algorithm in terms of
> its ability to remove artifacts and perspective 2 would be the performance
> of the removal algorithm in terms of its ability to NOT remove desired EEG
> signals.
>
> The method I had in mind was to identify artifact epochs, and then
> calculate the correlation coef between the reference channels and EEG
> channels before and after artifact removal (for perspective 1) a lower
> correlation should indicating better artifact removal.  Perspective 2 would
> be similar but instead would compare between unmodified EEG channels and
> channels with artifacts removed except that it would be during epochs
> without identified artifacts.
>
> In pursuing this, I am wondering if EEG lab has a function or plugin that
> can identify artifacts epochs given a reference channel.
>
> Second question; does anyone have any recommendations for better ways to do
> this?
>
> Thanks,
> CJ
> _______________________________________________
> 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
>