Related Experiment Videos
The polar average reference effect: a bias in estimating the head surface integral in EEG recording
M Junghöfer1, T Elbert, D M Tucker
1Department of Psychology, University of Konstanz, Germany. markus.junghoefer@uni-konstanz.de
Summary
The average reference in multichannel electroencephalography (EEG) approximates an inactive reference but requires sufficient electrode coverage. Inadequate or uneven electrode placement can introduce bias, known as the polar average reference effect (PARE).
Area of Science:
- Neuroscience
- Biophysics
- Electrical Engineering
Background:
- A reference-independent measure of potential is crucial for multichannel electroencephalography (EEG) analysis.
- The average reference, calculated as the mean of all recording channels, serves as an approximation of an inactive reference.
- This approximation is contingent upon accurate spatial sampling of scalp electrical fields.
Purpose of the Study:
- To investigate the validity of the average reference in multichannel EEG.
- To identify and explain the factors contributing to inaccuracies in the average reference.
- To introduce and define the polar average reference effect (PARE).
Main Methods:
- Simulations and analysis of EEG data with varying electrode densities and distributions.
- Evaluation of the average reference's accuracy under different spatial sampling conditions.
- Demonstration of the polar average reference effect (PARE) using illustrative examples.
Main Results:
- The average reference is a valid approximation of an inactive reference only with sufficient electrode density and full head coverage.
- Inadequate electrode density or uneven distribution leads to a biased average reference, termed the polar average reference effect (PARE).
- The PARE introduces smaller potential differences in central regions and larger differences at the periphery of electrode clusters.
Conclusions:
- Accurate spatial sampling is essential for the reliable use of the average reference in EEG.
- The polar average reference effect (PARE) can significantly distort EEG potential measurements.
- Maximizing electrode coverage of the volume conductor improves the accuracy of the average reference, bringing it closer to an ideal inactive reference.