Page 1 of 1

Post hoc measurement of EEG quality

Posted: Thu Jan 07, 2021 5:58 pm
by Douwehorsthuis
Good morning,
I was wondering if there is some sort of way to say anything about the quality of EEG data after a recording is done.
I can come up with simple ways like:
1) amount of channels that have too much high frequency noise
2) channels that have too much drift
3) channels that have are flat

But is there also a way to say anything about data that doesn't have such obvious problems. For example, is there a way to see (both during collection and afterwards) if channels are bridged? Is there a way to see if the electrode was maybe to far from the scalp (by adding a lot of gel we tend to overcome issues like this, but I am slightly worried that this might cause other issues)? Is the offset of channels stored anywhere, I assume this could maybe show some insightful info? Is there something I am missing that might be insightful for this?

Thanks for your time,
Douwe

Re: Post hoc measurement of EEG quality

Posted: Thu Jan 07, 2021 10:54 pm
by Coen
Offsets are readily extracted from the data, see http://www.biosemi.nl/forum/viewtopic.php?t=56

Best regards, Coen (BioSemi)

Re: Post hoc measurement of EEG quality

Posted: Mon Jan 11, 2021 9:09 pm
by Douwehorsthuis
Thanks for the offset info, but it's unfortunately not entirely clear how to see this when importing this data in matlab (EEGLAB). Am I correct to say that it is (the amplitude for each sample point) - (the average of 20sec of data)?

Besides offsets, are there other ways of saying anything about the data? Is there a way to see if channels are bridged and is there a way to see if the electrode was maybe to far from the scalp (by adding a lot of gel we tend to overcome issues like this, but I am slightly worried that this might cause other issues)?

Thank you,
Douwe

Re: Post hoc measurement of EEG quality

Posted: Tue Jan 19, 2021 2:44 pm
by Coen
Offset is just the average of a few seconds data (or low-passed data with a cut-off frequency < 1 sec)

Inject just enough gel to fill the holder, but not more (or you may introduce gel bridges). Use (rotate) the tip of the syringe to gentle work the gel between the hair onto the skin. The electrode itself does not touch the skin, the gel in the holder should provide the current path between electrode tip and skin.

Several methods have been published to extract gel bridges from the data, for example: https://www.sciencedirect.com/science/a ... via%3Dihub

Best regards, Coen (BioSemi)