EEG and ECG recording

This forum deals with BCI2000 configuration issues.
Locked
Elizabeth
Posts: 19
Joined: 20 Jan 2005, 16:29

EEG and ECG recording

Post by Elizabeth » 20 Jan 2005, 17:20

Hello,

I am currently setting up a BCI experiment where I will have subjects hooked up to EEG and ECG electrodes. Although the ECG data will not be used for the BCI application, I want to put it through the same amplifier and record it alongside the EEG data so that it will be precisely time synced to the EEG data. However, I was wondering if I could use the BCI2000 filtering on the ECG data? Do you know if this has been done before and if it sounds feasible to you?

Also, I assume that if I do use BCI2000 to filter my ECG data, I would have to include the ECG channels in the TransmitCh list and then put in an appropriate value for the SourceChGain. Do you have any suggestions for determining and selecting an appropriate gain?

Thanks,
Elizabeth

gschalk
Posts: 615
Joined: 28 Jan 2003, 12:37

Re: EEG and ECG recording

Post by gschalk » 21 Jan 2005, 17:27

I am currently setting up a BCI experiment where I will have subjects hooked up to EEG and ECG electrodes. Although the ECG data will not be used for the BCI application, I want to put it through the same amplifier and record it alongside the EEG data so that it will be precisely time synced to the EEG data. However, I was wondering if I could use the BCI2000 filtering on the ECG data? Do you know if this has been done before and if it sounds feasible to you?
Please remember that the filtering is all done solely for the purpose of real-time feedback. In other words, no filtering within BCI2000 will affect what is stored in the data file. Thus, unless you would like to control the system with somebody's heart beat (which could be fun, actually ;-)), this won't be of any use.
Also, I assume that if I do use BCI2000 to filter my ECG data, I would have to include the ECG channels in the TransmitCh list and then put in an appropriate value for the SourceChGain. Do you have any suggestions for determining and selecting an appropriate gain?
The gain/offset values are determined by your data acquisition system. The gain basically tells you how the conversion factor for each channel to convert the units that the data acquisition system produces into micro Volts. You can record a sine wave of a known amplitude, store it, and then look at the sine wave offline to determine the conversion factor. The GenerateCalibration program does this automatically for you (if you feed a sine wave to all channels simultaneously).

Gerv

Elizabeth
Posts: 19
Joined: 20 Jan 2005, 16:29

Re: EEG and ECG recording

Post by Elizabeth » 25 Jan 2005, 15:05

The gain/offset values are determined by your data acquisition system. The gain basically tells you how the conversion factor for each channel to convert the units that the data acquisition system produces into micro Volts. You can record a sine wave of a known amplitude, store it, and then look at the sine wave offline to determine the conversion factor. The GenerateCalibration program does this automatically for you (if you feed a sine wave to all channels simultaneously).
Hello,

I tried using the Generate Calibration program for the purpose above, but got some questionable results. Do you know what types of gain values should be expected when using a TDT system amplifier and a data translation board for EEG recordings? I tried putting both a 1V and 10V amplitude sine wave in and used the calibration program to get the SourceCh Gain / Offset values. I received drastically different values when using 1V vs. 10V. (I did change the peak to peak of input wave in the program from 100 (default) to 1000000 and 10000000 microV, respectively.) The values also differed based on whether I used the Screening_DTADC.bat or RJB_DTADC.bat program to record and store the data. Is this typically how you would store the data for this purpose... just by running one of the programs for a short period with a sine wave input to all channels?

Please let me know if you have any suggestions and if you know in what ballpark the gain and offset values typically are with a setup similar to ours.

I really appreciate all of the help you have given us so far!

Thanks,
Elizabeth
UW-Madison

gschalk
Posts: 615
Joined: 28 Jan 2003, 12:37

EEG recording ...

Post by gschalk » 25 Jan 2005, 19:50

Elisabeth,

Yes, you are correct in that you need to record a little piece of data and then run this through the GenerateCalibration program.

In theory, you do not need this program however. You should - in fact, you need to - first understand what happens to the sine wave that you put through your system. The 1V that you put in will correspond to some voltage that is input to the BCI2000 Data Translation Board. You can either measure this voltage using an oscilloscope, or you can derive it from studying the TDT system specs. Once you know the voltage that is input to the DT board, you can figure out what the conversion to muV should approximately be. (The DT Board digitizes at 12 bit=+-2048). Thus, the whole input range of the DT board (+-5V) is mapped to a range of +-2048 A/D units.

Once you have a ballpark idea what the gain (which is essentially the conversion factor that translate DT A/D units into microVolts) should be, then you are in the position to use the GenerateCalib program to get exact values.

You do you necessarily need any of these values. You could just use the same small values for each channel, e.g., 0.003. However, if you do not understand what happens to the signal, it is difficult to figure out if there are problems. You might consult an EE to help you with this.

Gerv

Locked

Who is online

Users browsing this forum: No registered users and 2 guests