We have observed a large noise increase in facial EMG recording when skin conductance electrodes (GSR) are applied. This seems to be caused by the fact that GSR amplifier produces currents that are captured by fEMG sensors. This is a serious problem because this noise (of about 200 microVolts) is usually stronger than the electrical signal from facial muscles. I wonder whether you might provide me some feedback whether you encounterd the same problem while simmulatenous EMG and SC recordings, and how you managed to handle it. We use ADInstruments Powerlab with BioAmp and GSR Amp. Thank you!
If you can do impedance measurements at a high frequency rather than galvanic response at low frequency or DC, then you will be able to filter it easily from your EMG signal. EMG is typically well below 500Hz and it should be reliable (and possibly more accurate) to measure tissue impedance above 5kHz. Measuring at high frequencies eliminates a large portion of the motion noise artifacts from the measurement.
If it is not possible to switch to higher frequencies then you can turn it around at lower frequencies. There is rarely any EMG signal with frequencies under about 5Hz so if you can do DC or low frequency GSR, you can run a 5Hz long pass on the EMG signal.
As mentioned before, your signals should occupy different bands of the frequency spectrum (surface EMG: 50-500Hz; GSR:
A wild guess is everything I can give. Extending the idea by Brian (above) maybe you could measure GSR on an extremity such as hand and wire patient to ground somewhere between the measurement site and the thorax. The GSR response is local as I guess that it is done by a standard four electrode method and grounding should set the skin potential to zero but I'm not sure if this helps. It could also help to decrease the GSR current. If it is not possible with the equipment, than you could put a resistor in parallel with the patient.
- Badges
- Science topic
- Similar topics
- Medicine
- Physiology