bad signals?

hannahhannah
edited November 2017 in Cyton
Hi,
I have a cyton board with 8 electrodes. I have been trying to record some EEG data but I have some issues with the signals. First, I struggle to get the signals from all the electrodes (is it the case for some of you?). But also, when I try to treat/analyze the differents channels it is challenging because the signals are so noisy that I have to get rid of a lot of trials (sometimes the whole session). Those images are examples of some of my filtered data:

image
image
image
I have some mouvement artefact too everytime I ask the people to make a motor response (squeeze a dynamometer).
image

The best data I can observe is this (for instance):
image

Do some of you also observe noisy data like these ones? I don't really know what to do with it, should I just spend more time processing it or should I do something about the possible interference that I can have when I record? Is it related to the fact that I use dry electrodes? Are gel electrodes better?

Thank you for your help!
Hannah
 
 

Comments

  • wjcroftwjcroft Mount Shasta, CA
    Hannah, hi.

    Paste / cup or gel electrodes will give more consistent signals. With your dry setup are you using the Ultracortex?

    When you connect all channels plus reference plus Bias, it should result in a signal close to 0 microvolts. Can be done with a blob of Ten20 paste or clip leads.

    Your first graph shows strong ~ 100 uV signals around 12 hz. Where could these be coming from? Are the site / channel labels correct? 

    Try to reduce as much as possible any cable motion, head motion, scalp tension, etc.

    William
  • hannahhannah
    Thank you for your reply.

    I use the Ultracortex Mark 4 EEG headset. The labels of the channels are correct. I don't know where the 100 uV signals may come from. Could it be movement artefacts?

    I'll try to connect all channels, reference and bias together to see what I can get.

  • hannahhannah
    I connected each channel at a time to the bias and the reference with some alligator clips. For all channels I got in average 0.13 mV (measured with the openBCI gui), is that fine? Thanks.
  • wjcroftwjcroft Mount Shasta, CA
    I think you mean microvolts, not millivolts (uV vs mV). Yes close to zero is expected. This means the amplifier is probably not your noise source. Can you try experimenting with gold cup / paste, or pre-gelled electrodes?

  • hannahhannah
    Yes I meant uV. I can't try experimenting with wet electrodes, I only have the dry electrodes that come with the ultracortex.
  • hannahhannah
    I think I am going to test some wet electrodes. I saw that sintered Ag|AgCl electrodes may be of good quality. Do you have perhaps any recommendations?
  • wjcroftwjcroft Mount Shasta, CA
    Although sintered silver chloride is one of the best signal quality disk or cup electrodes, they can be challenging to maintain.


    The gold cups sold by OpenBCI work just fine and are inexpensive. Ten20 paste is standard for attachment.
  • hannahhannah
    Thanks again. And can I use any of thoses electrodes with the cyton board? (gold cup or sintered silver chloride?)
  • wjcroftwjcroft Mount Shasta, CA
    Yes.
  • hannahhannah
    Ok great!
  • wjcroftwjcroft Mount Shasta, CA
    EEG amplifiers such as Cyton and Ganglion will work with all electrode types. This includes various metals (silver, silver chloride, gold, tin), physical shape (cup, disk, comb). And pre-gelled foam adhesive silver chloride; the foam types work best on bare skin.
Sign In or Register to comment.

© Vanilla Keystone Theme 2026