Eye-blinks: positive or negative?
PedroMen
Lisbon
I used Cyton board to acquire some EEG data mainly from frontal sites. The OpenBCI GUI presents eye-blinks as negative following positive peak. But I've seen in publications representations of only positive peaks or only negative peaks (valleys). What is the correct representation?
Is this correct?
Any links to publications that explain what is an eye-blink ?
Comments
Pedro, see some of these past Forum threads on EOG electrooculography.
https://www.google.com/search?as_q=eog&as_sitesearch=openbci.com
EOG is a form of EMG, with electrode positions that specifically pick up eye movements and eye blinks. The pick up of these signals and the form they take, depends on where the EEG / EOG / EMG electrodes are placed.
https://www.google.com/search?q=electrooculography
William
Thank you William but I still have doubts. My connections were:
Since Fp1 and Fp2 are connected to the N pins (the lowest, more close to the board), it means that the signal is inverted, right?
Yes, others have reported that their ERP waves are inverted, because OpenBCI uses the INxN pins as channel inputs instead of the INxP pins.
So reversing the polarity in your software is not going to make the pulses 'only' negative or positive. The only way to absolutely control the polarities is with the EOG electrode placements on either sides of the eyes.
But since is only about blinks (not left/right eye movements), these peaks should therefore be valleys (negative), right?
The only way to absolutely control the polarities is with the EOG electrode placements on either sides of the eyes.
Here is a good video on the subject:
Thanks for the video @Billh, but that video is just ridiculous: it explains nothing. Nevertheless, youtube suggestions lead me to some helpful explanations.
William, I wrote a specific montage, with specific electrodes, mounted specifically on the OpenBCI Cyton board, and you could not tell me if the output was correct????
Anyway, I found the answers to my questions:
Therefore, when there is an eye blink, Fp1/Fp2 will register a positive charge with reference to ear lobe. These positive values are represented (by convention) as a valley not as a peak.
On the GUI time series graph, positive Y axis values ARE positive, and negative are negative. However we already mentioned that the INxN pins used for channel inputs are the negative / minus inputs to the ADS1299 differential amps. The INxP pins, positive are bussed together as your SRB2 pin. Thus the 'normal' sense of graph is reversed. However this is only true for Cyton, not for other amps.
https://www.bem.fi/book/28/28.htm
The potentials around the eye depend heavily on the location of your electrodes. EOG montage is most accurate. Placing your reference on the EAR LOBE is fine for EEG, but for measuring eye potentials, I am thinking much less accurate since it is so far from eyes. There is no 'neutral' electrode potential. As the differential amps are always doing subtraction.
Thus eye artifacts which are not measured with an EOG montage, are very likely going to differ between individuals depending on their particular physiology, dimensions, conductivity, etc.
Is that really true? I know the video mentions this, but if there is any motion like this it must be extremely subtle. If my gaze is fixed on a point and I blink, I sense no eye movement when eyes reopen.
As far as I know, Yes, it's called Bell's phenomenon. Here are some scientific references:
From reference [1]:: Encyclopedia of the Neurological Sciences (Second Edition) - 2014, Pages 405-408:
"Scottish anatomist and surgeon Sir Charles Bell (1774–1842) is most often remembered for his descriptions of peripheral facial paresis (Bell's palsy) and the synkinetic upward rotation of the eyes with eye closure (Bell's phenomenon)"
From reference [2]:: Fundamentals of Sleep Medicine - 2012, Pages 545-566:
References:
1) https://www.sciencedirect.com/science/article/pii/B9780123851574008265
2) https://www.sciencedirect.com/science/article/pii/B9781437703269000270
So there we have it again, ear lobe is not "near the eyes". So differential amplifiers are going to respond more unpredictably with distant references.
The eyeball, as a rotating semispherical 3D object, is relatively symmetrical in 2 of 3 of certain 3D axes. Therefore it can be portrayed as an electrically polarized dipole with the pupil of the eye at one end relative to the back of the eye (the retina) as the other end of the electrically polarized dipole.
When the eyes blink closed, the pupil of the eye rolls up by reflex (this is likely part of a protective reflex in case the eyeblink is due to something moving toward the eye). This moves one end of the dipole toward the frontal EEG electrodes. As a result, a rather large EEG wave deflection occurs, followed by return to baseline at end of blinking as eyes open and the pupil part of the eyeball rolls downward to center again.
Do you need further explanation?