Extract features from an EEG (single channel) signal

Hi, I'm trying to extract features from an EEG; I've not too much experience in Matlab, so I don't really know how to proceed. Could someone help?

[Aakash Deep]

What do you mean by features. Are they muscular/ocular artifacts which you want to extract? Can you elaborate on this a little?

[Chiara Parrillo]

I need to extract alfa and theta rhythm to train a classifier

[Sandro Lecci]

You would want to extract the frequency components of your signal using either Fourier transforms or Wavelet approaches. This is a good start.

Best, Sandro

Answer 1

[Aakash Deep]

Hello Chiara,

If you want to extract the alpha and theta rhythm of the signal, you can find the wigner-ville transform of the signal (from which you can easily see the frequency band of the signal at any time point). This can actually help. A built-in function of this is not available in MATLAB but there are toolboxes which are supported in MATLAB. Below are two toolboxes for this:

TFTB: http://tftb.nongnu.org/

HOSA: https://in.mathworks.com/matlabcentral/fileexchange/3013-hosa-higher-order-spectral-analysis-toolbox

NOTE: MathWorks does not guarantee or warrant to use the content of these submissions. Any questions, issues, or complaints should be directed to the contributing author.

Hope this helps :)