English
 
Help Privacy Policy Disclaimer
  Advanced SearchBrowse

Item

ITEM ACTIONSEXPORT

Released

Conference Paper

REG-ICA: A new hybrid method for EOG artifact rejection

MPS-Authors
There are no MPG-Authors in the publication available
External Resource
No external resources are shared
Fulltext (restricted access)
There are currently no full texts shared for your IP range.
Fulltext (public)
There are no public fulltexts stored in PuRe
Supplementary Material (public)
There is no public supplementary material available
Citation

Klados, M., Papadelis, C. L., & Bamidis, P. D. (2009). REG-ICA: A new hybrid method for EOG artifact rejection. In Proceedings of the 9th International Conference on Information Technology and Applications in Biomedicine. doi:10.1109/ITAB.2009.5394295.


Cite as: https://hdl.handle.net/11858/00-001M-0000-0024-C5EC-6
Abstract
The plethora of artifact rejection (AR) techniques proposed for removing electrooculographic (EOG) artifacts from electroencephalographic (EEG) signals can be separated into two main categories. The first category is composed of regression - based methods, while the second one consists of blind source separation (BSS) - methods. A major disadvantage of BSS-based methodology is that the artifactual components include also neural activity, thus their rejection leads to the distortion of the underlying cerebral activity. The current study tries to solve the aforementioned problem by proposing a new hybrid algorithm for EOG AR. According to this automatic approach, called REG-ICA, independent component analysis (ICA) is used to decompose EEG signals into spatial independent components (ICs). Then an adaptive filter, based on a stable Version of the recursive least square (sRLS) algorithm, is applied to ICs so as to remove only EOG artifacts and maintain the neural signals intact. Then the cleaned ICs are projected back, reconstructing the artifact - free EEG signals. In order to evaluate the performance of the proposed technique, REG-ICA has been compared with the least mean square (LMS) approach, in simulated EEG data. Two criteria were used for the comparison: how successfully algorithms remove eye blinking artifacts, and how much the EEG signals are distorted. Results support the argument that REG-ICA removes successfully EOG activity, while it minimizes the distortion of the underlying cerebral activity in contrast to LMS.