Localization of Brain Activity from EEG/MEG Using MV-PURE Framework

TL;DR

This paper introduces a new reduced-rank spatial filtering method based on the MV-PURE framework for more accurate brain activity localization from EEG/MEG data, especially in challenging conditions.

Contribution

The paper develops novel reduced-rank activity indices with proven unbiasedness and improved spatial resolution over traditional methods for EEG/MEG source localization.

Findings

Higher spatial resolution in localizing closely positioned sources

Effective in low signal-to-noise ratio conditions

Validated with simulated and real data examples

Abstract

We consider the problem of localization of sources of brain electrical activity from electroencephalographic (EEG) and magnetoencephalographic (MEG) measurements using spatial filtering techniques. We propose novel reduced-rank activity indices based on the minimum-variance pseudo-unbiased reduced-rank estimation (MV-PURE) framework. The main results of this paper establish the key unbiasedness property of the proposed indices and their higher spatial resolution compared with full-rank indices in challenging task of localizing closely positioned and possibly highly correlated sources, especially in low signal-to-noise regime. Numerical examples are provided to illustrate the practical applicability of the proposed activity indices using both simulated and real data.