Isolating gait-related movement artifacts in electroencephalography during human walking
Julia E Kline1, Helen J Huang, Kristine L Snyder
1Department of Biomedical Engineering, University of Michigan Ann Arbor, MI, USA.
Journal of Neural Engineering
|June 18, 2015
Summary
High-density electroencephalography (EEG) during walking is contaminated by movement artifact not explained by head acceleration. Current artifact removal methods are insufficient, necessitating advanced techniques for accurate brain activity analysis.
Area of Science:
- Neuroscience
- Biomedical Engineering
- Signal Processing
Background:
- High-density electroencephalography (EEG) offers insights into brain function during real-world activities like walking.
- Quantifying movement artifact versus electrocortical activity in walking EEG data remains challenging.
Purpose of the Study:
- To characterize movement artifact in EEG during walking across various speeds.
- To evaluate the effectiveness of artifact removal techniques.
- To compare EEG movement artifact with head acceleration.
Main Methods:
- A novel method isolated movement artifact using a nonconductive swim cap and a conductive wig.
- EEG data were recorded from nine subjects walking at speeds from 0.4 to 1.6 m/s.
- Moving average and wavelet-based artifact removal methods were tested.
Main Results:
- Movement artifact varied significantly with speed, subject, and electrode location.
- EEG movement artifact showed poor correlation with head acceleration.
- Tested artifact removal methods reduced low-frequency noise but did not eliminate artifact; spectral patterns mimicked published walking EEG data.
Conclusions:
- EEG data during walking contain substantial movement artifact, independent of head acceleration.
- Artifact characteristics vary, and traditional signal processing methods are inadequate for removal.
- Development of more sophisticated artifact removal methods is crucial for advancing EEG research in locomotion.
