The Study Of Artifact Removal Based On EEG

The Electroencephalograph (EEG) is the biological signals that reflect the electrophysiologicalactivity of brain cells. It is produced by a large number of neurons cells, and carries lots ofphysiology and pathologic information. It can be applied to the field of brain-computer interface,etc. The potential of EEG signal is very weak, and mixed with a lot of artifacts (such as eyemovment, heartbeat, muscle jiggle and power frequency, etc.). These artifacts have brought greatdifficulties to analyze and use EEG, so the artifact removal from EEG becomes especiallyimportant.First, this paper introduces the classification and characteristics of EEG, describes many kindsof artifacts and recent artifact removal methods. The theories and methodologies of principalcomponent analysis (PCA) and independent component analysis (ICA) have been discussed. Thenthe paper introduces a relative new removal method——Hilbert-Huang Transform (HHT).Empirical Mode Decomposition (EMD) is the critical part of HHT. Aiming at the end effect inEMD, this paper proposes a new method which combines mirror extension with waveformmatching extension. The simulation verifies the superiority of the improved Hilbert-HuangTransform. The HHT has been used on real EEG signal, according to IMF’s power spectrum andother information, we can successfully identify and remove some artifacts from EEG.Because the HHT method is hard to remove eye movement and electrocardiogram(ECG), thispaper, finally, provides a new method: use HHT to filter the signal, and then use PCA to decreasethe dimension of the signal, finally, use ICA to remove the artifacts. Use this method on real EEGwe can remove high frequency, eye movement and ECG from EEG. Summarizing advantage anddisadvantage of this method, the paper discusses the future development of EEG artifact removalmethods.