Saturated EEG Channels

[darvish_sam]

Hi everyone,
I craeted a batch file based on the comments in this forum for a Refa8 TMSi 64 channel EEG amlifier and it worked.
However, when I followed the instruction for a Mu-session, It looks like that all eight channels ( F3, F4, C3, Cz, C4, T7, T8, Pz)have been saturated since they showed very big values like "-39492223.999999" and no signal is being showed in the signal view page.
BTW , I checked the channels impedance by the ANT companion software( ASA) and there were all between 10 - 20 Kohms, I also checked the signals with ASA , and they were again big signals ( peak to peak ~ 1000 muv), however, after adding notch filter( 50 Hz)and 5 Hz low pass and 30 Hz high pass filtering, it could be shown like a normal signal ( peak to peak < 100muv)

Surprisingly, when I add the same notch filter and high pass filter in BCI2000 invironment, (using right click on the signal view page) there was no changes in the signals, but after adding 5 Hz low pass filter, all the signals gradually decreased to 0.000 !!

Could you please let me know what could be wrong?

Thanx,
Sam

[boulay]

Do you mean 5 Hz high pass and 30 Hz low pass?
5Hz high pass will quickly bring your signals to have a mean around 0. Are you suggesting that it is completely flat at 0.000...?

I can't help you with your problem, but I just wanted to help you clarify your question so someone more familiar with the TMSi system can better help you.

[darvish_sam]

Do you mean 5 Hz high pass and 30 Hz low pass?
5Hz high pass will quickly bring your signals to have a mean around 0. Are you suggesting that it is completely flat at 0.000...?

I can't help you with your problem, but I just wanted to help you clarify your question so someone more familiar with the TMSi system can better help you.

Thank you for your correction.
Yes, actually it was a typo in my question and I mean the signals were filtered so as to only show the frequencies between 5-30 Hz.
As for being flat at 0.000, yes it gets completely flat graduallay ( in few seconds).

Thanks,
Sam

[mellinger]

Hi,

the problem might be unrelated to BCI2000. Do you get reasonable EEG signals when using the software that came with the amplifier?

Large offsets can be due to wrong cabling, or to mixing different electrode materials. Make sure that both the amplifier's reference and ground input are connected to electrodes which are attached to the head, and filled with gel. Also make sure that all electrodes are made from the same electrode material. Otherwise, electrochemical potentials will build up between electrodes, which will saturate the amplifier.

HTH,
Juergen

[darvish_sam]

Hi,

the problem might be unrelated to BCI2000. Do you get reasonable EEG signals when using the software that came with the amplifier?

Large offsets can be due to wrong cabling, or to mixing different electrode materials. Make sure that both the amplifier's reference and ground input are connected to electrodes which are attached to the head, and filled with gel. Also make sure that all electrodes are made from the same electrode material. Otherwise, electrochemical potentials will build up between electrodes, which will saturate the amplifier.

HTH,
Juergen

Dear Juergen,
Thank you for your reply. As you mentioned the problem did not relate to BCI2000. The thing was that I had to change the connection betweenthe 8 channels and EEG amlifier.
To clarify it more for the folks who in the future want to connect Refa ANT amplifiers and BCI2000, I connected F3 channel to channel Number 1 on EEG amlifier, F4 to 2, T7 to 3, C3 to 4,Cz to 5, C4 to 6, T8 to 7 ,and Pz to 8. That way BCI 2000 could receive the signals correctly.

Thanks,
Sam

[ruudkalis]

Hi, this is Ruud from ANT.

The Refa ANT amplifier is a so called DC amplifier. That means that it measures all signals from DC to abour 1/3 of the sampling frequency. There is only a low-pass filter at 0,27*Fs in the amplifier (for anti-aliasing purposes). There is no high-pass filter, which explains why there is DC-offsets in your signal. To get "normal" values, please filter the data before further processing. More information on "why" DC amplifiers are used can be found in the following article:
http://www.bmedreport.com/archives/3739

[darvish_sam]

Hi, this is Ruud from ANT.

The Refa ANT amplifier is a so called DC amplifier. That means that it measures all signals from DC to abour 1/3 of the sampling frequency. There is only a low-pass filter at 0,27*Fs in the amplifier (for anti-aliasing purposes). There is no high-pass filter, which explains why there is DC-offsets in your signal. To get "normal" values, please filter the data before further processing. More information on "why" DC amplifiers are used can be found in the following article:
http_nospam_www.bmedreport.com/archives/3739

Hi Rudd,
Thanks to your clarification, the above mentioned problem was solved , but now I am trying to use the same amplifier ( TMSi-Refa8 with 64 channels) for online classification of motor imagery EEG signals. Since our defined classification update rate is fast, using BCI2000 with windows XP gives me long roundtrip warning with simulated EEG data but it works fine with windows 7 64 bit.
Thus I am wondering if it is possible to install the mentioned EEG amplifier in windows 7 64 bit. And if it is possible where I can find the required drivers for that?

Thank you in advance,
Sam