Neurofeedback calibration

[protoconsciousness]

Hi again,

I am setting up a new neurofeedback scenario very similar to the 'Neurofeedback' example scenario.
Because every subject would have different baselines, I need the mean and variance of a previous calibration session. ("This two boxes are configured with respect to a previous calibration session. The calibration computed the Mean and Variance of the Beta activity during 30sec of inactivity." as the 'Neurofeedback' example puts it)

How do I set up a calibration session, extract Mean and Variance and feed this values into the neurofeedback scenario?

Thank you very much!

Best,

Clemens

[fabien.lotte]

Hello,

to setup a baseline scenario, you could use the lua scripting box to generate stimulations indicating the beginning and end of the baseline, and indicate to the user these stimulations with a sound (sound player box) or a picture for instance (display cue image box). Then you record/extract the EEG signals between these two stimulations (e.g., with a Stimulation based epoching box) and compute the mean variance using the univariate statistics box.

I hope this helps.

Best regards,
Fabien

[protoconsciousness]

Hi!
Thanks a lot for your help!

Using the univariate statistics box to get the mean and the variance, is there a way to automatically write those values into a simple DSP or a crop box? For the mean for example I want to have a DSP box calculates 'x - mean' (x is the signal). Would I have to start a calibration scenario, read out the mean, open the feedback scenario and type the value in manually each time or is there an easier way?

Thank you again!

Clemens

[jtlindgren]

One way to hack this is to write the value to a configuration file in the calibration scenario which produces the mean. The config file is then read by the Simple DSP box in the next scenario with the 'override settings...'

You can also use configuration tokens (e.g. ${Blah}) in the simple DSP formula, but then the problem is to set the token to the mean value, and this is not necessarily easier to do (usually these tokens are read from .conf files).


Happy hacking,
Jussi