5 Lead ECG

BioPotentialBioPotential
edited April 2019 in Cyton
Does OpenBCI support recording ECG signals from a 5 lead ECG cable?

I have a few spare ECG cables that are 5 leads but have 6 wires terminated into a header. The sixth wire seems to be shield.
AFAIK EEG signals can have a single reference but ECG signals are measured differently with various references.

My leads are RA (right arm), RL (right leg), LA (left arm), LL (left leg), and V (chest lead) and possibly 1 shield.

I was thinking of connecting the RA, LA, LL and V to the the Positive
inputs of the Cyton and  connect RL to bias (considering that this would
be Right Leg Drive).

Would it work this way?
What do I consider as reference(SRB1/SRB2) and where do I connect the shield, should I drive the shield with the bias signal?.

Do I need to do some special programming to get a 5 lead ECG to work with Cyton / OpenBCI?

Regards,

Comments

  • wjcroftwjcroft Mount Shasta, CA
    edited April 2019
    Bio, hi.

    Here is a related thread,


    But not actually that relevant to 5 lead.

    I would suggest just ditching your 5 lead cables that have a header-plug intended for mating with some type of medical equipment. Getting a reliable connection with that medical equipment plug could be problematic.

    Instead just use the touchproof to header pin adapter that is sold in the shop. That way you can mate with any ECG snap electrode wires. These are inexpensive, only about $15 for the set of 5 snap leads.


    image

    ----

    As far as hooking up 5 lead. As you say, ECG is entirely differential. Does not use SRB2 or 1. So in the GUI settings (or your own app using the SDK), you will want to turn off SRB on those channels. You will need 3 channels for the 3 "Leads" of 5 lead ECG (numbered Lead I, Lead II, and Lead III). See the 5 lead diagram in this document, on page 2,


    image

    There are 3 derived 'Leads', and equations for those, I find different definitions, you'll have to decide which ones are correct. See Goldberger's equations in these docs, the first looks simpler,


    Then this,


    ----

    Hookup.

    You need to make 3 Y-cables. These cables will have 2 female header connectors spliced to 1 male header connector. The wire length can be short just a couple inches. These ribbon packages can be separated and cut, then spliced and wrapped with tape or shrink tube.


    The wiring uses the Y cables to bridge pins on the three channels, as shown. [I FINALLY re-edited this post figure to have the correct polarities on all Leads.] The male pin from the Y, goes to your touchproof converter.


    image

    As far as the Bias, yes that goes to RL, it's sometimes called DRL Driven Right Leg. As far as I can tell, the V electrode is not used(?)

    Regards,

    William
  • BioPotentialBioPotential
    edited April 2019
    Amazing and mind blowing. This is too good!
    Many many many thanks William!

    I see you have wired up the Leads in reverse polarity to what is standardly done.
    Correct? Because AFAIK
    Lead I = RA- and LA+
    Lead II = RA- and LL+
    Lead III = LA- and LL+

    Correct?

    So assuming we wire it up in either you way and the standard way we will get the 3 waveforms in the OpenBCI GUI.
    So how does one calculate the Augmented leads. Does one have to use the SDK for that (i mean based on the equations you provided in the given link)? Or is there some way to set it up in the GUI?

    What would be the general procedure to calculate the Augmented waveforms. I guess they would have to be done in real time as the data is is streaming is? Are they just a simple summing of the 2 channels? And is it possible in OpenBCI GUI to sum channels up that way?
     
    PS: The "ecgwaves" link is enligtenlingly amazing. Many thanks!
  • wjcroftwjcroft Mount Shasta, CA
    edited April 2019
    Bio, hi.

    re: polarities reversed (again!)

    Will post a revised diagram.

    re: calculation of augmented leads

    Have you seen the OpenBCI_Python repo? With this you can use DSP digital signal processing libraries to receive the samples one at a time and then apply various filters or calculation expressions. Python also has excellent output graphing libraries. As I understand the Goldberger's equations they involve simple math, but not just "summing channels".

    No, the GUI cannot do expression evaluation.

    Another way you could go is with a VPL, such as BrainBay, OpenViBE or even National Instruments LabView. All these have expression evaluator elements.

    Regards,

  • wjcroftwjcroft Mount Shasta, CA
    OK, 'FINAL' edit on the ECG wiring diagram, see the first reply above. Apologies for my polarity confusion. Was looking at the arrows in the "Standard 5-Lead Placement" diagram and not thinking.

  • BioPotentialBioPotential
    edited May 2019
    These are the first test results. Of Lead I and Lead II only.
    For some reason I feel the waveforms look inverted.But I've connected as the diagram given above.
    Is the top row of the the input connector the IN*P pins when looking from the side where the ADS1299 and MCU are mounted, that is top view of the board.

    image



    image
  • wjcroftwjcroft Mount Shasta, CA
    On the Cyton, the default IN*N pins are the row of pins CLOSEST to the board surface. The IN*P pins are farthest pins from the top of the board.

    On the Settings tab view, I would not 'include' Bias to any of the channels. This is mostly used for EEG to inject inverse common mode cancelation. I'm not sure it is relevant to ECG. Since the ECG is in millivolts, and the EEG microvolts.

    Bias was attached to the RL position?

    I would not use auto-scale on the GUI. Instead select a manual scale that works for ECG level signals. Turn off other channels that are not being used.
  • BioPotentialBioPotential
    edited May 2019
    I was playing around with the filter settings and when I selected "No Notch" and "No BP" then the waveforms no longer showed on screen. I tried all ranges but still no luck. I had to switch to auto to get it back on screen.

    Regarding the IN*N and IN*P pins the wiring seems to be correct based on what you describe, I have connected the bias to the farthest pin. But for the bias I guess both the pins in the same column are shorted.

    Bias is used in any biopotential measuring device for reducing the effect of common mode interference that the patients body picks up. Please have a look at https://en.wikipedia.org/wiki/Driven_right_leg_circuit
    It's amazing that even with the right leg drive I see lots of 50Hz interference if I disable the Notch and BP filters. The bias was meant to correct this. I wonder if we can tune the bias accurately to null of the 50Hz pickup.

    And yes I connected the bias electrode to the right leg.

  • wjcroftwjcroft Mount Shasta, CA
    No, from all I've seen with the ADS1299 chip, the counter mains (common mode) injection is not strong enough to cancel out the mains noise. You always need a notch. Other ECG equipment might be different. But all EEG equipment I've seen uses some types of mains DSP filtering, either bandpass or notch.

    'Including' Bias in only ONE of the channels, is not recommended. Either use it in NONE, or in all your channels. Unfortunately with all the noise on you other channels, you want to definitely turn those off with the colored buttons.

    Some DRL circuits might be tuned specifically for ECG, perhaps clinical equipment. ADS1299 is general purpose, primarily EEG. An EEG inverse mains injection would likely be not the same as ECG inverse mains injection. EEG operates in microvolt scales, ECG in millivolt.

    I still suggest you turn off the auto scale. If you have to turn on a bandpass filter to do that, please do. Remember the recording is unfiltered so all your raw data is there.
    With the bandpass off, you might be seeing artifacts from the DC offset, typically in the 10s of millivolts.
  • BioPotentialBioPotential
    edited May 2019
    I have quite long ECG 5 Lead cables, that may be contributing to the extra noise on the channels. Also they have a shield wire. But I guess there is no connection for shield on the Cyton.
    1. Also is there a way of measuring the CMRR of the whole system(electrode cables + cyton)  with the electrodes connected?
    2. Does Cyton + Open BCI GUI allow to measure the electrode impedance? I guess the ADS1299 has a feature that allows for meaurement of lead impedances.
  • wjcroftwjcroft Mount Shasta, CA
    On the GUI time series panel, the impedance is measured with the 'ohm' symbol, next to the colored channel on/off button. The ohm symbol is a toggle switch. When on, it reports the impedance per channel.

    Generally with EEG and likely ECG as well, the shorter the leads the better. Inexpensive 5 ECG snap lead sets can be had in a variety of lengths.



    Actually shield wires sometimes affect overall capacitance of your inputs. Especially if they are not connected. The obvious suggestion for connection of your shield wires is to connect them with the ground lead that is present on the mains wall connector. This mains ground is NOT present on any of the OpenBCI board pins.

    There are some other threads on CMRR measurements, use the search button in the right column.

Sign In or Register to comment.

© Vanilla Keystone Theme 2026