Signal Acquisition for P300 speller

[Sona Balsam]

We, the engineering students are doing a project on 'Thought Commands Using Non Invasive BCI' using the OpenVibe software.
We are trying out the demo of P300 speller for practice.
Is there a way to give pre recorded EEG signals as input to the P300 speller rather than attaching any acquisition device?
If there is one, can you please tell the procedure?

[ddvlamin]

You can use one of the boxes in the 'File reading and writing'-section and connect the appropriate inputs and outputs. However, for the P300, you will have to make sure that the triggers in your data correspond to the ones you set in your 'Stimulation based epoching'-box so that the correct epochs are averaged.

[yrenard]

Dear Sona and Dieter,

another option is to use the scenarios which filename include the -replay tag. Their purpose is exactly to replay files that were recorded online.

I hope this helps,
Yann

[vaishnavi.j]

Dear Dieter,
In the snapshot that you have posted as reply, only part of the implementation is visible.
Can you post a detailed implementation flow chart of p300 speller?
In case you have any pre recorded p300 signal, can you post it as attachment?

[ddvlamin]

Hi,

As Yann says you can use the default scenario files that come with openvibe, for replaying recorded data check the scenario file with the "-replay" suffix.

I can upload two p300 data sets, one for training and one for testing: you can find them here: http://www.thewired.be/blog/wp-content/ ... v_p300.zip

I think there is also a prerecorded p300 signal in the share\openvibe-scenarios\signals directory that comes with openvibe.

First you will need to train the xdawn spatial filter and then the classifier with their corresponding scenario files. Furthermore I used 12 repetitions. The channels with index 12 and 18 should not be included, so you should select all other channels by means of the channel selector box.

Best regards,
Dieter Devlaminck