how to measure electrode impedance?

edited November 2016 in Hardware
(apologies in advance if this should go in Hardware or if another thread already -- please merge if so)

Is there a way to measure the impedance of the electrode to skin contact, via the OpenBCI board and/or software? Or is the only method to disconnect the electrode and reference wires from the board and use an ohmmeter to manually test it?

What do you guys do to make sure you're getting low impedance between electrode and scalp? I have the dry, spiky electrodes from the kit ... and wonder if they're actually making good electrical contact.

Comments

  • Great! Thank you!
  • edited November 2016
    Quick question about the paper you posted in the discussion 84 link (http://www.csi.uoregon.edu/members/ferree/pubs/EEG2001.pdf)

    It says "the required procedure was to abrade the skin to achieve a scalp-electrode impedance of less than 5 kΩ. To achieve such impedance levels, skin abrasion is required.  Abrasion removes the surface epidermal layer, which has higher impedance than the underlying tissue."

    If that's true, then how am I getting such low impedances with just the dry, spiky electrodes used in the headset?? I'm not even using gel.


    And then perhaps a follow up comment: while I'm no expect, I am not so sure that I agree that just because you have low impedance, means you're getting accurate EEG measurements. Perhaps that statement is wrong to the point of being silly. But to my way of thinking, low impedance just means the electrode conductors of the channel and the reference have good electrical contact with the scalp/ear skin, and thus are well able to pass current along the surface of the skin.

    Why am I even saying this? Because I wonder if the geometry of the spiky electrode is in fact "limiting" in its ability to detect neuronal (EEG) activity. Yes, great impedance (gets through hair, without mess of gel), but does that in fact mean good EEG detection??

    Pretty much just wondering aloud. But I do think the geometry matters. There is a guy at U of Rhode Island who has great results using "Tri-Polar Concentric Ring" electrodes. Anyway, rambling now ...
  • wjcroftwjcroft Mount Shasta, CA
    Leka, hi.

    re: (sometimes) low impedance of dry combs

    Impedance varies with the pressure applied. Since the dry combs are usually applied with a headset or velcro band, there is considerably MORE pressure on the skin than with cups and paste. On the other hand, impedance with the dry combs or cones can vary more for this same reason. Impedance of gel or paste electrodes is generally more constant during a recording session; because the applied pressure or contact is more uniform.

    The geometry of the combs DOES help reduce impedance because it digs deeper into the skin. 

    re: combs "limiting its ability to detect EEG". No, good conductivity is all that is required. Another geometry used by Cognionics is a comb with flexible tines that bend and splay outward as pressure is applied. OpenBCI has some reels of very low resistance conductive printer filament and will be trying some experiments with that. There are pictures of that and more discussion over on the Dry Electrodes thread in Electrodes category.

    William
  • "No, good conductivity is all that is required."

    That doesn't seem right, to me. I doubt that the neurons are directly supplying charges to the conductors, through the skull. If that were the case, then what you said would make sense. Rather, I think the EEG sensors work more like how TV antennas work. And in that sense, the shape of the antenna does affect reception.

    Conductivity between the sensor conductor and the scalp is meaningful for impedance, obviously.
  • Thanks for the link to VC. It seems to confirm what I thought, that EEG working first principles are fields, not direct electricity. That makes sense to me.
  • wjcroftwjcroft Mount Shasta, CA
    Yes, the electrodes pickup ensemble activity from large numbers of neurons. Like holding a microphone up to the edge of a football stadium.  :-)  Each individual neuron firing is considered a "dipole" in the VC model.

  • And so, I wonder just how good of a "stadium microphone" the spiky dry electrodes are, compared to the flat rings (using gel to remove air/hair gap).

    Or like I mentioned before, the guy at URI uses concentric rings to obtain even higher performance.
  • wjcroftwjcroft Mount Shasta, CA
    I've seen papers comparing different electrode metal types, for example Ag-AgCl, gold, tin, silver, sintered Ag-AgCl, etc. And other papers comparing active (built in amp inside electrode) vs. passive. The comb structure seems very common in dry schemes. As it reaches through the hair. I believe Cognionics has several papers on their site comparing their dry comb headset to gel approaches.

    http://www.cognionics.com/
  • wjcroftwjcroft Mount Shasta, CA
    re: "It seems to confirm what I thought, that EEG working first principles are fields, not direct electricity."

    Yes, at the scalp you are measuring ensemble electric (or magnetic in the case of MEG) combined field effects which result in an electric field (micro voltage) potential difference between the electrode being measured and the reference electrode.

    But this is not like the static (high voltage) electric fields measured in physics lab experiments. Where the electrode shape concentrates charge. Nor is it like RF radio frequency electromagnetics, where again the antenna shape effects the received signal strength.

    By the time the summed charges reach the scalp, it's a simple micro voltage potential difference. So flat or comb or gel type shapes don't show any appreciable differences.
  • wjcroftwjcroft Mount Shasta, CA
    edited November 2016
    Here's the pdf of the Tri-Polar concentric ring electrodes,

    http://egr.uri.edu/wp-uploads/neurorehabilitationlab/IEEE-TBME-vol53-no5-pp926-933-2006Besio-et-al.pdf

    And the Rhode Island PhD thesis using them,

    http://digitalcommons.uri.edu/oa_diss/69/

    ----

    I'm still unclear how the three rings in the electrodes (PC board apparently) are connected to amplifier channels. I think one conductor may act as a reference and the other two conductors are two separate channels that are then mathematically combined as in the paper.


  • wjcroftwjcroft Mount Shasta, CA
    Found a photo of the tri-polar ring electrodes. These are ACTIVE devices with a circuit board on the back of the electrode. See the 3rd page of this paper,


  • edited November 2016
    Thanks for the links on the tri-polar. Active "just" means that the amplifier is sitting on top of the electrode, instead of being back at the board/box, right? (I put just in quotations, because that's not trivial or cheap, I'm sure)

    I've only been to one talk by Besio, so that's how I knew of them. I guess he was at LA Tech before. Anyhow ...


    "By the time the summed charges reach the scalp, it's a simple micro voltage potential difference. So flat or comb or gel type shapes don't show any appreciable differences."

    I don't see why the relatively low strength of the summed electric fields from the action potential currents means that you can completely ignore electrode shape, and just throw any old piece of conductor on your scalp. And even if that were the case, then what's the point of insuring low impedance between the electrode conductor and the scalp? I guess the larger the air gap, the larger the distance between the source of the field and the electrode, and thus the smaller the signal strength. But that also means because of the spikes on the dry comb, the bulk of the electrode conductor is sitting off the scalp surface ...

    Appreciate any further comments.
  • wjcroftwjcroft Mount Shasta, CA
    The Cognionics papers are here,

    http://www.cognionics.com/index.php/technology/download-now

    Some of them show comparison graphs and such with dry combs vs. gel / paste systems. They conclude no significant differences.

    Spikes on the combs are directly digging into the skin layers. So distance of the comb framework / backing makes no difference. There are a range of papers published on the electrical and mechanical characteristics of the skin / electrode interface.
  • Appreciate your comments (and links especially), as always. Thank you
  • Hi...i am a new user here. As per my knowledge the dry combs are usually applied with a headset or velcro band, there is
    considerably MORE pressure on the skin than with cups and paste. On the
    other hand, impedance with the dry combs or cones can vary more for
    this same reason. Impedance of gel or paste electrodes is generally more
    constant during a recording session.
Sign In or Register to comment.