Cyton + Ultracortex checking / calibration suggestions ? [resolved]

RodRod Bristol. UK
edited May 2021 in Cyton

I'm wondering there is a way of checking that the electrodes are working properly (on the 8 channel mark IV headset). Is there a way of calibrating the result / plot from the channels?
I have my mark IV headset running with the OpenBCI GUI and through Brainflow (python) directly.
The plots from the GUI look as attached.
Also I guess I am asking because I have been trying to use the Brainflow EEG Metrics example which uses an ML model to read concentration and relaxation
https://brainflow.readthedocs.io/en/stable/Examples.html#python-eeg-metrics
and this seemed to work well a couple of days ago and now does not...

If anyone has any ideas I would be hugely grateful!

Comments

  • RodRod Bristol. UK

    I also wondered what 'Railed' and 'Not Railed' means...

  • wjcroftwjcroft Mount Shasta, CA

    Rod, hi.

    You are in the UK, so you should be using the 50 Hz notch filter. It looks from your FFT, that the 50 Hz peak is huge, swamping out all other valid EEG. Press the 'Notch' button to the right of the red 'Stop Data Stream' button. That should flip the notch from 60 to 50, and return your valid EEG to more normal range.

    If you have questions on the Brainflow metrics, it's best to ask on the Brainflow Slack channel. My understanding is that this ability is still in beta testing and further development.

    https://brainflow.org/ [see Slack link]

    Regards, William

  • RodRod Bristol. UK

    Hi William, thanks! as ever! Giving this a try now.
    I'm on the brainflow slack group, so will try to port this info over.

  • wjcroftwjcroft Mount Shasta, CA

    An easier check than the metrics you mention, is the increase in alpha with the eyes closed.

    https://docs.openbci.com/docs/01GettingStarted/02-Biosensing-Setups/EEGSetup#4-alpha-brain-waves-eeg

  • RodRod Bristol. UK

    Why does is make a difference that I am in the UK?

  • wjcroftwjcroft Mount Shasta, CA

    The electricity grid supplied from power plants, runs at different frequencies, depending on your location. It's 60Hz here in the US and I believe in Japan. Much of the rest of the world runs on 50Hz.

  • RodRod Bristol. UK

    Hi William, thanks! That really seemed to make a difference, trying the blink test now.

  • wjcroftwjcroft Mount Shasta, CA

    You can do the eyes closed test with a recording, then play it back. Or with a friend. Alpha is stronger in the rear of the head (occipital and parietal lobes.)

  • wjcroftwjcroft Mount Shasta, CA

    If you are seeing 'railed' on channels 5 and 6, it usually means poor contact with scalp. You can check impedance while you are wiggling around the combs on the head. Also see this thread.

    https://openbci.com/forum/index.php?p=/discussion/1481/how-to-become-de-railed

    https://www.google.com/search?as_q=railed&as_sitesearch=openbci.com

  • RodRod Bristol. UK

    Is the interference from the electricity grid affecting my brainwaves? Or their translation in the GUI / software?
    i.e is the interference in me or the machine?

  • wjcroftwjcroft Mount Shasta, CA

    I'm not sure I understand your question. When you are using the wrong notch filter, yes the huge 50 Hz component is swamping out the real EEG signal. However with the channel 5 and 6 'railed', that is unrelated to the mains noise. Instead you have an issue with less then optimal scalp contact with those electrodes. See the links.

  • RodRod Bristol. UK
    edited May 2021

    Hi William. I guess I am asking where the 'swamping' is happening. Is it literally the local grid affecting my brainwaves, or is that it creates an artefact in the software because (maybe) the computer is wired to the local electricity grid.
    If it is the former it is interesting that the electricity grid is affecting my brainwaves!

  • RodRod Bristol. UK

    Thanks for the tip on the railed electrodes too. I will try to fix that. Oddly they did all seem to have contact with my scalp.

  • wjcroftwjcroft Mount Shasta, CA

    I guess I am asking where the 'swamping' is happening. Is it literally the local grid affecting my brainwaves, or is that it creates an artefact in the software because (maybe) the computer is wired to the local electricity grid.

    Yes, with the wrong notch filter selected, the local mains grid is so huge compared to microvolt brain signals, that it is literally 'everywhere' in the environment. It has nothing to do with software artifacts, or computer plug. However the mains interference is STRONGEST, near wires and devices that have a mains connection: extension cords, power cords, power supplies, LED and CFL lights, floor / wall / ceiling conduits, etc. The 'field' of mains intensity is strongest within a few feet of these sources. As you get farther it tapers off quickly.

    With the railed channels, check your impedance, you will find likely those channels are much different than your normal channels.

  • wjcroftwjcroft Mount Shasta, CA

    Is it literally the local grid affecting my brainwaves...

    It's not "affecting your brainwaves", instead it is swamping the microvolt level sensitivity of the analog to digital converter on the Cyton. This converter does not know how to distinguish your brainwaves from mains interference.

  • RodRod Bristol. UK

    Thanks William!!

  • RodRod Bristol. UK

    Hi William, A quick update. I worked out which electrodes were railing and swapped them out for the two spares that came with the 8 channel kit and all is good with no railing!
    Thanks for your help!

Sign In or Register to comment.