Weighted phase lag index stability as an artifact resistant measure to detect cognitive EEG activity during locomotion

Troy M Lau,Daniel P Ferris,Kaleb G McDowell,Joseph T Gwin

doi:10.1186/1743-0003-9-47

Abstract

BackgroundHigh-density electroencephalography (EEG) with active electrodes allows for monitoring of electrocortical dynamics during human walking but movement artifacts have the potential to dominate the signal. One potential method for recovering cognitive brain dynamics in the presence of gait-related artifact is the Weighted Phase Lag Index.MethodsWe tested the ability of Weighted Phase Lag Index to recover event-related potentials during locomotion. Weighted Phase Lag Index is a functional connectivity measure that quantified how consistently 90° (or 270°) phase ‘lagging’ one EEG signal was compared to another. 248-channel EEG was recorded as eight subjects performed a visual oddball discrimination and response task during standing and walking (0.8 or 1.2 m/s) on a treadmill.ResultsApplying Weighted Phase Lag Index across channels we were able to recover a p300-like cognitive response during walking. This response was similar to the classic amplitude-based p300 we also recovered during standing. We also showed that the Weighted Phase Lag Index detects more complex and variable activity patterns than traditional voltage-amplitude measures. This variability makes it challenging to compare brain activity over time and across subjects. In contrast, a statistical metric of the index’s variability, calculated over a moving time window, provided a more generalized measure of behavior. Weighted Phase Lag Index Stability returned a peak change of 1.8% + −0.5% from baseline for the walking case and 3.9% + −1.3% for the standing case.ConclusionsThese findings suggest that both Weighted Phase Lag Index and Weighted Phase Lag Index Stability have potential for the on-line analysis of cognitive dynamics within EEG during human movement. The latter may be more useful from extracting general principles of neural behavior across subjects and conditions.

Highlights

High-density electroencephalography (EEG) with active electrodes allows for monitoring of electrocortical dynamics during human walking but movement artifacts have the potential to dominate the signal
Movement artifacts in EEG recorded during walking include movement of electrodes, loss of skin contact, muscles activation associated with head stabilization [6,7], and cable sway that leads to electronic interference
We investigated a measure of Weighted Phase Lag Index (WPLI) Stability (WPLIS) to generalize the WPLI measure, classifying periods of high or low stability/volatility

Summary

Introduction

High-density electroencephalography (EEG) with active electrodes allows for monitoring of electrocortical dynamics during human walking but movement artifacts have the potential to dominate the signal. One of the most fundamental and difficult aspects of this challenge is to parse EEG from electromyographic, electroocular, and movement artifacts that occur during movement [1,2,3,4,5] Overcoming this challenge would help researchers understand the cognitive dynamics that occur during everyday life. Other electrical artifacts in EEG occur due to muscle activation associated with jaw clenching and blinking, and movement of the eye (electroocular artifacts). These latter artifacts are not specific to movement tasks but can still make it difficult to separate out EEG related from true cognitive dynamics

Methods

Results

Conclusion