Unified Bayesian estimator of EEG reference at infinity: rREST

TL;DR

This paper introduces a unified Bayesian framework for EEG referencing, combining average reference and REST into a single model, and demonstrates improved estimation accuracy through simulations and real EEG analysis.

Contribution

It formulates EEG reference estimation as a Bayesian inverse problem, unifying AR and REST, and develops regularized estimators with superior performance.

Findings

rREST achieves lowest potential estimation error in simulations

Real EEG analysis shows rREST outperforms traditional methods

Volume conductor models enhance REST and rREST performance

Abstract

The choice of reference for electroencephalogram (EEG) is a long-lasting unsolved issue resulting in inconsistent usages and endless debates. Currently, both average reference (AR) and reference electrode standardization technique (REST) are two primary, irreconcilable contenders. We propose a theoretical framework to resolve this issue by formulating both a) estimation of potentials at infinity, and, b) determination of the reference, as a unified Bayesian linear inverse problem. We find that AR and REST are very particular cases of this unified framework: AR results from biophysically non-informative prior; while REST utilizes the prior of EEG generative model. We develop the regularized versions of AR and REST, named rAR, and rREST, respectively. Both depend on a regularization parameter that is the noise to signal ratio. Traditional and new estimators are evaluated with this framework, by both simulations and analysis of real EEGs. Generated artificial EEGs, show that relative error in estimating the EEG potentials at infinity is lowest for rREST. It also reveals that realistic volume conductor models improve the performances of REST and rREST. For practical applications, it is shown that average lead field gives the results comparable to the individual lead field. Finally, it is shown that the selection of the regularization parameter with Generalized Cross-Validation (GCV) is close to the 'oracle' choice based on the ground truth. When evaluated with the real 89 resting state EEGs, rREST consistently yields the lowest GCV. This study provides a novel perspective on the EEG reference problem by means of a unified inverse solution framework. It may allow additional principled theoretical formulations and numerical evaluation of performance.