Page 1 of 1

reference without EEG

Posted: Tue Jun 05, 2012 4:51 pm
by KaPo
Hi,
I have a question concerning the reference: I exclusively record EMG (zygomaticus Ex1+2, corrugator Ex3+4), GSR, Plet and Respiration.
For analyzing the data, do I have to build the difference between Ex1 and 2, and accordingly, between Ex3 and 4 in order to get a bipolar EMG signal? Additionally, we have large pulse artefacts here. Is it due to the position of the CMS?
And how is the GSR signal referenced?
Are all signals already rectified and integrated?

Best regards,
Katrin

Re: reference without EEG

Posted: Wed Jun 06, 2012 4:58 pm
by Coen
Calculating bipolar leads by subtracting EX1-EX3, EX3-EX4, etc. is the correct approach.

If the heart is in between the CM and the EMG electrodes, the pulse artifact may be ECG. Such an artefact is Common Mode for the EMG leads, and is suppressed with a factor of 40-60 dB after subtracting. The Common Mode arteact itself can be decreased by locating the CMS electrode in the vicinity of the EMG electrodes.

The GSR is calculated form the voltage drop between the GSR electrodes, the reference plays no role.

The signals are clearly not rectified and integrated (original higher frequency EMG waves are displayed). Rectifying and low-pass filtering may of course be applied during replay (or on-line by adding a bit of LabVIEW code).

Best regards, Coen (BioSemi)