COMPEL: The International Journal for Computation and Mathematics in Electrical and Electronic Engineering

Purpose – The purpose of this paper is to evaluate the performances of a resolution scheme able to follow the dynamics of brain tissue properties in combined ElectroEncefaloGraphic (EEG) – MagnetoEncefaloGraphic (MEG) techniques for the brain analysis, minimizing the computation burden.

Design/methodology/approach – The estimation process in combined EEG-MEG is performed by a Moore-Penrose pseudo-inverse computation. This is affected by the uncertain knowledge of the living tissues' electric properties. In principle, it is possible to estimate those properties from the EEG-MEG signals. The estimation process becomes in this case non-linear. A resolution scheme is proposed, based on the exploitation of the different dynamics characterizing sources and tissues properties.

Findings – The proposed resolution scheme provides a reasonable estimate of the sources for a computationally affordable frequency of non-liner estimations.

Research limitations/implications – The proposed approach has not been tested yet on experimental data, and as such, its sensitivity to environmental uncertainty is not known yet.

Practical implications – The proposed strategy can be easily implemented to perform realistic measurement processing.

Originality/value – The paper presents a novel strategy to estimate tissues properties and EEG-MEG signal sources based on the exploitation of their different dynamics, possibly taking advantages from an impedance tomography preliminary analysis for the tissue properties dynamics.