Delivering electric shocks while recording EEG

Post Reply
Noa
Posts: 2
Joined: Fri May 24, 2019 5:43 pm

Delivering electric shocks while recording EEG

Post by Noa » Fri May 24, 2019 5:58 pm

Hi,

We have a new experiment that involves delivering electric shocks (using Biopac stimulator) while recording EEG with BioSemi. We have noticed that while the stimulator is on, the EEG data looks clean, same as when the stimulator is turned off. However, when we attach the two electrodes from the stimulator to the subject’s arm (without even delivering a shock), there is a very robust high-frequency noise. We have a few questions with regard to that:
1. The lowpass filter in the ActiView “monopolar display” screen is set to 100 Hz. If we change this to 20 Hz, the signal looks much better. Just to make sure that we understand correctly – Does that mean that the raw data is digitized using an online lowpass filter?
2. It is possible to change the cutoff value during the experiment? Meaning, to use 20 Hz in blocks were subjects are connected to the stimulator, but to return to the 100 Hz in blocks where they are not.
3. Do you have any other suggestions of how to rearrange the setup, in order to minimize this electric noise?

Any help would be highly appreciated.
Thanks in advance!

Coen
Site Admin
Posts: 965
Joined: Fri Mar 26, 2004 7:00 pm
Location: Amsterdam, Netherlands
Contact:

Re: Delivering electric shocks while recording EEG

Post by Coen » Mon May 27, 2019 12:43 pm

Apparently, the simulator jeopardizes ActiveTwo's perfect galvanic isolation provided by battery power and fiber-optic data transfer. This causes high Common Mode (50/60 Hz interference on all channels). Reject Common Mode by referencing. The simulator is probably mains powered or connected to mains powered equipment. Correct this situation to reduce Common Mode and improve subject safety (provide proper galvanic isolation for the simulator setup).

The filter settings on the left sidebar only affect the signals on screen, the data to BDF file is always full bandwidth (1/5th of the sample rate). You should reference and filter data offline.

Best regards, Coen (BioSemi)

Noa
Posts: 2
Joined: Fri May 24, 2019 5:43 pm

Re: Delivering electric shocks while recording EEG

Post by Noa » Sun Jun 02, 2019 12:55 pm

Thanks!!

We will check the galvanic isolation issue with Biopac.
During offline processing, we regularly re-reference the data to the average of the two electrodes placed on the mastoids. This does not eliminate the noise in this case...

Do you think there is anything at all we could do online (in addition to offline), that would help reducing this interference, in terms of the BioSemi's settings?

Thanks again!

Coen
Site Admin
Posts: 965
Joined: Fri Mar 26, 2004 7:00 pm
Location: Amsterdam, Netherlands
Contact:

Re: Delivering electric shocks while recording EEG

Post by Coen » Mon Jun 03, 2019 12:39 pm

The software is deliberately designed to only save the unfiltered (bandwidth: DC to 1/5th of the sample rate) and unreferenced data.

Best regards, Coen (BioSemi)

Post Reply