[Methods for Processing Physiological Artifacts in Single/Few-Channel EEG Signals]

Guojing Wang,Hongyun Liu,Weidong Wang,Hongyan Kang
DOI: https://doi.org/10.12455/j.issn.1671-7104.230374
2024-05-30
Abstract:Electroencephalogram (EEG) is a non-invasive measurement method of brain electrical activity. In recent years, single/few-channel EEG has been used more and more, but various types of physiological artifacts seriously affect the analysis and wide application of single/few-channel EEG. In this paper, the regression and filtering methods, decomposition methods, blind source separation methods and machine learning methods involved in the various physiological artifacts in single/few-channel EEG are reviewed. According to the characteristics of single/few-channel EEG signals, hybrid EEG artifact removal methods for different scenarios are analyzed and summarized, mainly including single-artifact/multi-artifact scenes and online/offline scenes. In addition, the methods and metrics for validating the performance of the algorithm on semi-simulated and real EEG data are also reviewed. Finally, the development trend of single/few-channel EEG application and physiological artifact processing is briefly described.
What problem does this paper attempt to address?