Signal Quality Changing over Time: CM rejection problem?

[jonatmudd]

Hi all,

Seems everyone here is using biosemi for EEG/MEG.

I use the biosemi to measure intestinal activity, same concept as measuring brain activity due to phsyiological currents.

We use the biosemi ActiveTwo.

My problem is that the signal quality degrades over an experiment of about several hours.

During the first ~1 hr or so, everything is rational and we are able to record signal from the intestine we expect to see.

However, as the experiment progresses, we always experience that this intestinal signal is no longer recorded. Instead we measure a signal at the respiration rate of the test subject on all electrode channels. By about 1-2 hrs, we are no longer able to measure any of the intestinal signal, and measure only this respiration on all channels.

This smells like a grounding issue or a common mode rejection problem, but can't figure out what we are doing wrong. Hopefully some one out there can help.

Details below, for any interested party.

We place 16 channel Ag/AgCl electrode platform directly on the exterior (serosal wall) of the intestine of our test subject (anestheized pig). These are essentially surface electrodes that may slightly puncture the intestinal wall...at any rate they are in direct contact with the intestine.

Our grounding configuration is as follows.
The DRL electrode is connected to a stainless steel surgical needle which is stuck into the muscle of the right leg of the animal. The CMS electrode is a cutaneous electrode, attached the skin surface of the pig, centered over the intestines. Additionally, there is an electrode that is attached to the shielding of the 16 electrode platform that we (sometimes) connect to the DRL electrode. The quality of our signals seems to be the same whether this connection is made or not.

Any ideas why, over a time frame of a couple of hours, we see a common respiration signal on all channels? Any ideas on how to fix the problem?
Again, during the initial stages of the experiment, everything is fine, recordings are rational. It is only after some time that our measurements crap out.

Any help would be very much appreciated.

jon

[Coen]

I would initially suspect the DRL needle (stainless steel can be a problematic electrode material). I would recommend to use a normal surface electrode for the DRL (preferable Ag-AgCl). By the way, the DRL does not have to be located on the right leg, a location near the CMS is fine (but keep the distance between CMS and DRL large enough to prevent gel bridges).

Best regards, Coen (BioSemi)