Ganglion Setup Confusion / EEG + ECG
Hello OpenBCI Forum.
I've been setting up my ganglian with hopes of simultaneous ECG and EEG to monitor heart-brain effects, but to no complete avail.
As the tutorial for setup instructs, I have two ear electrodes connected to the upper D_G and REF pins and I've also got two frontal EEG electrodes connected to the upper pins of channel 2 and 3. Additionally, all switches are in the DOWN position.
My main question is: why is there a heartbeat like signal in situation when all electrodes are NOT connected to me, and also when the D_G and REF electrodes ARE connected while the frontal scalp EEG electrodes are NOT.
Weirdly enough, when I place the two EEG electrodes on my frontal scalp (electrode 2 on my left forehead and electrode 3 on my right), they both seem to be dominated by heartbeat like activity, whilst sometimes the signal dies down and seems to be more related to neural activity since I can get a blink response. But this is not consistent.
What is the likely reason for this?
Comments
Here is an image of some activity from the EEG electrodes when against my forehead.
Hi Austin,
Look over this recent thread about simultaneous EEG with strong signals such as EMG or ECG.
https://openbci.com/forum/index.php?p=/discussion/3414/two-pins-per-channel-simultaneous-eeg-with-emg-eog-or-ecg/
The EEG signal is microvolt level. The other biosignals are millivolt levels. So they tend to swamp out EEG, if using the same references. The solution is to switch ECG / EMG channels to differential mode, and use separate plus minus electrodes for those channels.
William
Also, in general, you cannot leave channel pins "floating" and not connected to skin or scalp. Because the amplifiers are so sensitive, when floating, the wires and pins act like antennas.
When the pins are left without a connected wire and I check their data it is pretty random which is to be expected. Though, when a wire is connected it is this heartbeat like signal showing up....what could this be?
And also, referring to your first message @wjcroft, should I thus have both reference pins and both ground pins connected?
an update: i still have heart-corrupted signal for the two forehead electrodes, even when i use both + and - reference and ground pins (i have both reference pins connected to my earlobes and both ground pins connected to my chest with ecg electrodes).
What to do? Is this a faulty device?
As mentioned previously, a wire connected to the Ganglion, but left floating and not connected to body, will pickup environmental noise.
No, I did not say that. Both REF pins are the same circuit connection. Both GND pins, are both the same ground.
Did you read the previous comments? Why are you connecting ground pins to the chest, these are likely why ECG is leaking into your EEG. Both REF pins are identical circuit connections. Both GND pins are identical GND connections.
No. If you looked at the previous comments / links, what is suggested is to place your switches in the common reference position for the EEG channels, and use one ear lobe for REF and the other for GND.
https://docs.openbci.com/Ganglion/GanglionSpecs/#inverting-input-select-switches
For your ECG channel, set its switch to the differential mode. And connect wires to the plus and minus pins on that channel. (To opposite sides of the chest.) For the ECG channel, no REF or GND is needed. This setup will eliminate the millivolt level ECG signal from leaking into the EEG channels.
Regards, William
Additionally, since the ECG signal is so much larger than EEG, do not bundle the EEG cables and ECG cables 'together'. Doing so may cause inductive or capacitive coupling between the ECG and EEG wires.
Thank you William. I have read the shared document and the past forum that you mentioned. Unfortunately I cannot get 2 good EEG channels running. I've removed the ECG wires just to try and at the least get good EEG.
I have one ear connected to ground and one connected to reference. I have all switches in the DOWN position. I have 2 EEG electrodes connected to my forehead. 1 channel yields meaningful signal and one does not.
Also, any idea what this peak at 20Hz could be? This peak is always there and is especially pronounced once every 10 seconds or so when the rest of the signals seem to die down.
Update: now they have switched, without me doing anything....
And just for clarity, this is the setup
What electrodes are you using? Dry or with paste? Suggest trying the cups with paste if you have that. Have you tried swapping cables / electrodes? It looks like you have the 50 Hz notch filter turned on? The 20 Hz may be a sub harmonic, or environmental noise. Try getting away from extension cords, floor / wall ceiling power conduits, electric lights, power supplies, metal cabinets / surfaces.
The 'so-called' heart like signal on channel 3, also has a huge amount of 20 Hz noise. Further evidence that this is somehow picking up environmental signals. A damaged electrode may be acting as reception point.
The "ECG-like" signal in all images is most likely not heart in origin, as the period is precisely 1.0 seconds. So this is coming from EMF (electromagnetic field) noise in your environment.
One of the dry electrodes you are using likely has a damaged or worn coating. Try swapping each component individually one at a time to isolate. Dry electrodes require MUCH more pressure from that headband. You could also try applying a bit of water moisture to the forehead spots before applying.
Interesting. Any idea what this 1 Hz EMF noise actually could be? I'm curious.
i'm using the dry comb electrodes as opposed to the flat ones. Would you expect the comb ones to work better?
I will buy the gold cup ones if this will make a substantial difference.
After some time the signals seem to have regulated themselves and are now more concentrated on neural activity. Though i didn't do anything for this to be achieved. Still concerning since it is not consistent.
The pulses are 1 Hz, but the background is intense 20 Hz, you can see that in the previous time series. Check your location and all the sources mentioned previously. This is environmental noise, some type of equipment in your area is radiating this. Get away from it.
The headband kit comes with flat electrodes for the forehead. That is what you should use.
If this happens again, adjust your location in the room or go to a different room. Also try swapping (one at a time) cables, electrodes, etc. A bad electrode could increase susceptibility to EMF noise.
Cups and paste or cap and gel or cap and saline sponges, will always perform better than dry electodes.
As mentioned, I believe one of your electrodes may be faulty. The dry electrodes have a plastic base, then plated with silver, then plated again with a chloride coating. If any of these plating layers wears down, the electrode will perform erratically.
William
Thanks William!
I've had a couple more sessions trying to figure this out.
It seems as though the order in which I plug wires into the ganglian is important. For example, when I start the data recording with all wires plugged in then the signal is inherently noisy (strong 20 Hz signal and repetitive 1Hz pulse). But when I start the recording without the ground and reference and then plug them in during the recording itself, the data seems to get regulated to, what I hope to be, meaningful EEG data.
Can you comment on this so that I know if something is faulty or if this is expected?
Also, can you verify that the comb electrodes are more suited for the scalp/hair whilst the flat ones are for the forehead?
Thanks for your help throughout, I really hope to get this up and running ASAP.
Have you tried moving your setup to another location? As mentioned the exact 1.000 Hz pulses are most likely coming from some equipment, cords, conduits, lighting, etc. in your immediate area. Same for the 20 Hz. Ground and Reference are always needed and should be plugged in along with all other electrodes at the session start. I can see how the sequencing you mention would affect this, unless some of your wires / electrodes are defective. Did you try swapping electrodes / cables as mentioned?
Combs are for hair area, while flats are for bare skin areas.