[Eeglablist] nonlinear ICA

[Vuckovic, Aleksandra]

Dear all,

 

I would like to use non-linear ICA on 21 channel recordings. I am working on detection of imaginary movements from single trials. I know that most people say EEG can be treated as a linear mixture but due to different conductivity through brain and skull and rest of tissues the resultant activity recorded by EEG cannot be really linear summation of sources.

 

Does anybody have some experience with non-linear ICA? I would like to try MISEP method but don't know if it works on some many channels?

 

If anybody have used non-linear ICA on EEG please recommend me a method and of possible the relevant literature.

 

Thanks a lot in advance!

 

Aleksandra Vuckovic

BCI group

University of Essex, UK