Connecting Actichamp with OpenVibe for Motor imagery

[sumitsoman]

We are using the Actichamp amplifier with the following electrodes connected: 8, 24, 25, 11, 12, 23, 22, 6, 7, 29, 28, and GND. These correspond to channels C3, Cz, C4, CP5, CP1, CP2, CP6, FC5, FC1, FC2, FC6.

We use OpenVibe Acquisition Server to acquire EEG from the amplifier, and view the EEG signal using a OpenVibe Designer scenario consisting of an Acquisition client and a signal visualization box. The EEG signals for the connected channels appear as straight lines, while for the other channels, we observed variations, as shown below (see channels 6,7,8).

Without connecting channel 1

Without_Channel_1_Connected.png (191.49 KiB) Viewed 10665 times

However, when we connect channel 1 (Fp1), there is a reversal in display of the EEG of the channels. In this case, the connected channels show variations whereas the other channels which are not connected appear as straight lines. This is shown in the figure below (see channels 6,7,8).

After connecting channel 1

With_Channel_1_Connected.png (169.66 KiB) Viewed 10665 times

What is the reason for this? Do the blank lines contain EEG signals?

[jlegeny]

Hello,

this could be a problem with the scale. Could you try to remove the non-connected channels from the display? You can do this either by using a channel selector box prior to the display or by configuring the signal display box and removing the unused channels from the list of channels.

Regards
Jozef

[lbonnet]

Hi everyone,

I guess this topic is the same as the one you can find on the BrainProducts actichamp forum :
http://www.actichamp.com/forum/viewtopic.php?f=4&t=94

For your interest, here is a duplicate of my answer there.

Hello cdac_bci,
I am Laurent Bonnet, engineer at Mensia Technologies.
I have developed the actiCHamp driver for OpenViBE, so maybe I can help you

To be sure of what we are dealing with, I have a couple of questions:
- What is the configuration of the openvibe driver, in the acquisition server ? Did you change any parameter ?
If you didn't change anything, I assume you have the default parameter, notably:
> Active shielding ON
> All available channel modules ON

On the hardware side, it's mandatory to always connect the channel 1 if you activate the shielding mode (or Impedance mode for the record). A good practice (according to BrainProducts) would be then to use a subset of electrodes in ascending order. You would need to put them on the right spots of the cap, without necessarily following the number tags.

As the shielding mode is, I guess, activated, this may explain the strange behavior when connecting or disconnecting the channel 1.

On the openvibe side, I have some comments too that may help.
The simple signal display box has some limitations. One feature is the auto-scale: the unique scale is chosen according to the channel that has the highest amplitude, which is good when dealing with a normal eeg setup. However, this feature is annoying when displaying unconnected (=pure noise) and connected (=eeg) channels at the same time, as their scales are completely different.

If you select only the connected channels, what do you see ?
To do so, add between the acquisition client box and the signal display box a Channel Selector box.
You can search for it in the search bar on top of the box tree-list.

Configure this box to select only the relevant channels (e.g. 6;7;29;28;8;24;25;11;12;23;22).
And tell me if you still see flat lines on these channels.

And the answer from original poster:

Thanks for the reply.

We have not made any changes to the parameters of the acquisition server. The Active Shield Gain is set to 5. For channel modules, only the button with label 1 (1-32) is active, and that is set to ON.

We find that the channel display shows all blank lines when we use the channel selector as well.

However, the reply given at http://www.actichamp.com/forum/viewtopic.php?f=4&t=93 suggests that it is mandatory to connect channel 1, as a result of which we are able to get the signals in the signal display window on running the OpenVibe scenario. The issue seems to have been resolved.

This problem seems to be solved

Laurent-