Research on EEG signal datasets collected by neurosky mindwave. Want to remove artifacts i.e, EOG, EMG.
I want to use Openvibe for this. Is anyone guide me for this or any other techniques to implement for removing such artifacts. It would be great help for me.
Why do you use Openvibe? You can use MATLAB without need for remove artifacts.
You might want to use Empiric Mode Decomposition (Sifting) for the signal which is part of the Hilbert Huang Transform in order to denoise with the use of weighted frequencies. However, I would not recommend to do such thing with the eSense reading from the Mindwave, since it is an already filterred value, you might want to use weighted average for your time series.
For OpenVibe, if you obtain the raw data for each electrode, you might want to denoise for each frequency band after the FFT (Spectogram and band power values)
Here is a good article about weighted frequencies
Article Seizure classification in EEG signals utilizing Hilbert-Huan...
I hope that this info is helpful
Here is one published approach:
J.L. Whitton, F.A. Lue and H. Moldofsky. "A Spectral Method for Removing Eye Movement Artifacts from the EEG" (EEG and Clin Neurophysiol 44:735-741, 1978).
If you are getting a unusually large amount of artifact in your EEG's, i suggest taking a review of your procedures. The signal/noise (s/n) of the differential amplifiers are highly dependent on their common mode rejection ratio. Furthermore the s/n for any one channel is dependent on the impedance of the electrode for that channel against the reference; so it is important to measure the electrode impedance.
Since EEG signal is the non-stationary signal and artefact is the transient contamination, therefore simple filtering like bandpass or lowpass filter will not work. The popular technique for removing EOG, EMG kind of artefacts are : adaptive Filtering, Wavelet Transform, and Empirical Mode Decomposition (EMD).
I think wavelet transform would be a better option for your problem. It is efficient due to its minimum signal distortion than EMD or another method. Wavelet transform with coiflets mother wavelet has been reported promising results in ocular and muscular artefacts elimination in recent studies in EEG artefact removal research.
- Badges
- Science topic