Frequency-based brain networks: From a multiplex framework to a full multilayer description

TL;DR

This paper investigates how representing brain networks as multilayer structures with inter-frequency interactions affects spectral properties, using synthetic and real MEG data to highlight the importance of interlayer edge heterogeneity.

Contribution

It introduces a detailed analysis of multilayer brain networks considering inter-frequency interactions, emphasizing the impact of interlayer heterogeneity on spectral characteristics.

Findings

Interlayer edge heterogeneity significantly influences spectral properties.

Full multilayer models differ from multiplex models in spectral behavior.

Real MEG data confirms the importance of considering inter-frequency interactions.

Abstract

We explore how to study dynamical interactions between brain regions using functional multilayer networks whose layers represent the different frequency bands at which a brain operates. Specifically, we investigate the consequences of considering the brain as a multilayer network in which all brain regions can interact with each other at different frequency bands, instead of as a multiplex network, in which interactions between different frequency bands are only allowed within each brain region and not between them. We study the second smallest eigenvalue of the combinatorial supra-Laplacian matrix of the multilayer network in detail, and we thereby show that the heterogeneity of interlayer edges and, especially, the fraction of missing edges crucially modify the spectral properties of the multilayer network. We illustrate our results with both synthetic network models and real data sets obtained from resting state magnetoencephalography. Our work demonstrates an important issue in the construction of frequency-based multilayer brain networks.