Comparing ERPs across OpenBCI, Biosemi, Emotiv

jessicalokejessicaloke
edited April 2018 in Research
Hello, I've recently completed a go/nogo task with OpenBCI. However, OpenBCI is not showing clear ERP's as other devices. I've used lab streaming layer to collect both EEG and Presentation trigger streams. Below is a plot showing the ERP's.

https://drive.google.com/open?id=1432VRvmKv7U4lprn1J7-1dHv3npwi0Fk

Here is the xdf file recorded with Lab Recorder from LSL. Could anyone help let me know what could be the issue?

Thank you in advanced.



Comments

  • wjcroftwjcroft Mount Shasta, CA
    Jessica, hi.

    All these three amplifiers use different electrode setups, for example the BioSemi uses active electrodes, hence the most accurate. 

    What electrodes are you using with OpenBCI? Does your EEG look good on the GUI screen? Any noise? Where are the references for these three systems, etc.? Have you experimented with different OpenBCI electrodes? The baseline kind of being gold cups with paste.

    William

  • wjcroftwjcroft Mount Shasta, CA
    Here are some of Jeremy Frey's measurements,


  • jessicalokejessicaloke
    I have the dry/spiky electrodes for my OpenBCI setup. On the GUI, it appears that some electrodes have better signal than others but I can still see clear artifacts and alpha waves. I've been using the electrodes for a few months now and I've noticed that the signals are getting weaker - thus, i also posted a question if I'm damaging the electrodes by using conductive gel on them and washing the gel off with hot water. Both Biosemi and Emotiv use references - CMS/DRL. And I only have experience with the dry electrodes because they are the only set I have. Thank you for your answers, William.
  • wjcroftwjcroft Mount Shasta, CA
    Are the reference and ground (Bias) leads in the same locations on each amplifier? Where on the 10-20 map? Ear lobes? 

    On the FRI electrodes, do you see a uniform grey coating (AgCl plating) on all surfaces? Do they look the same as when new? 
  • jessicalokejessicaloke
    Do you mean the reference for the other devices? For Emotiv it is on the bone right behind the ears. 
    The uniform grey coating looks fine, to the naked eye, it is as they are when they first came in.
  • wjcroftwjcroft Mount Shasta, CA
    What are the 10-20 locations used for OpenBCI? What electrode type are you using for reference and Bias? Have you tried moving these around, for example using the mastoid references, etc.? Can you list 10-20 locations for each amplifier.
  • jessicalokejessicaloke
    The 10-20 locations for 16 OpenBCI electrodes are the standard configuration provided in the tutorial. For reference and bias, they are connected with the ear clips provided. I've not moved them or deviated from the standard configuration for this dataset.
  • wjcroftwjcroft Mount Shasta, CA
    Are you using gel on the earclips also? If you use gel for the channels, the references should get it as well. 

    Try checking your impedances.

    A single earlobe reference may not be best for your ERP. Many neurofeedback clinical units use "linked ears", with left and right earlobes connected either by direct hardware, or software. Bias on the midline, Electro-cap uses AFz. Suggest looking at the Frey paper and see what he used.
  • jessicalokejessicaloke
    The impedence was <10 at the start of the experiment. I missed your question about the amplifier, for my experiment, I did not use an amplifier for both my mobile device - could that be the reason there were no clear ERPs then? 
  • wjcroftwjcroft Mount Shasta, CA
    Are you using gel on the earclips also? If you use gel for the channels, the references should get it as well.

    I don't understand your "I did not use an amplifier for both my mobile device" statement. By 'amplifier' I was referring to the three devices: OpenBCI, Emotiv, Biosemi. Each of these is considered an EEG amplifier. EEG signals are at microvolt levels and amplifiers record them.

    My only suggestion at this point is to try different electrodes. Get some gold cups and paste, cup electrodes are less than $10 each. Try the mastoid reference, etc.

    OpenBCI is capable of ERP P300 recording, as shown in the Frey paper. Unfortunately I have not done any P300 myself, so I'm just kind of giving you general suggestions.

    William


  • jessicalokejessicaloke
    Ah, thanks for the clarification. I misunderstood the word amplifier - thinking that the device is actually connected with an actual signal amplifier. I'll probably try different electrodes as an alternative. Thank you for the suggestions. Also, I'll try using gel for the earclips.
  • wjcroftwjcroft Mount Shasta, CA
    Is this a commercial gel product? Manufacturer / product name?

    All EEG amplifiers do contain "actual signal amplifer[s]", they are differential amps between + and - pins, or channel and reference.


  • jessicalokejessicaloke
    Yes, this is a commercial gel product - 'signa gel' made by Parker Laboratories. 
    Thanks again for the clarification.
Sign In or Register to comment.

© Vanilla Keystone Theme 2026