Information Optimized Multilayer Network Representation of High Density Electroencephalogram Recordings.

Francesc Font-Clos,Simone Sarasso,Maria Paola Canevini,Stefano Zapperi,Benedetta Spelta,Caterina A M La Porta,Armando D’Agostino,Francesco Donati

doi:10.3389/fnetp.2021.746118

Abstract

High-density electroencephalography (hd-EEG) provides an accessible indirect method to record spatio-temporal brain activity with potential for disease diagnosis and monitoring. Due to their highly multidimensional nature, extracting useful information from hd-EEG recordings is a complex task. Network representations have been shown to provide an intuitive picture of the spatial connectivity underlying an electroencephalogram recording, although some information is lost in the projection. Here, we propose a method to construct multilayer network representations of hd-EEG recordings that maximize their information content and test it on sleep data recorded in individuals with mental health issues. We perform a series of statistical measurements on the multilayer networks obtained from patients and control subjects and detect significant differences between the groups in clustering coefficient, betwenness centrality, average shortest path length and parieto occipital edge presence. In particular, patients with a mood disorder display a increased edge presence in the parieto-occipital region with respect to healthy control subjects, indicating a highly correlated electrical activity in that region of the brain. We also show that multilayer networks at constant edge density perform better, since most network properties are correlated with the edge density itself which can act as a confounding factor. Our results show that it is possible to stratify patients through statistical measurements on a multilayer network representation of hd-EEG recordings. The analysis reveals that individuals with mental health issues display strongly correlated signals in the parieto-occipital region. Our methodology could be useful as a visualization and analysis tool for hd-EEG recordings in a variety of pathological conditions.

Highlights

Recent developments in neuroscience are giving rise to an increasing amount of data on the functioning of the brain at different scales, from molecular processes at the level of single neurons to macroscopic signals encompassing the whole brain, as in electroencephalogram (EEG) or functional magnetic resonance imaging
We propose a multilayer network representation of EEG signals that maximize the information content and apply it to a set of sleep EEG data from patients diagnosed with First Episode Psychosis (FEP) or Bipolar Disorder (BD) and compared with control subjects
Network representations have been shown in the past to provide a useful tool to highlight the connectivity and spatio-temporal correlation of brain activity as revealed from EEG or other measurments such as functional magnetic resonance imaging (fMRI)

Summary

Introduction

Recent developments in neuroscience are giving rise to an increasing amount of data on the functioning of the brain at different scales, from molecular processes at the level of single neurons to macroscopic signals encompassing the whole brain, as in electroencephalogram (EEG) or functional magnetic resonance imaging (fMRI). A important goal is to develop tools that are able to extract useful information from brain activity measurements on individual subjects in order to identify potential network dysfunction and support diagnosis (Bassett, 2021). It is becoming increasingly clear that brain activity is strongly interconnected and hierarchically organized, requiring a sophisticated mathematical description to infer its underlying properties from measurements. The emerging field of network neuroscience is advocating the use of networks descriptions for a statistical analysis of brain functions at multiple spatio-temporal scales (Bassett and Sporns, 2017). A typical feature of many complex networks that appears promising to describe the hierarchical brain organization is the small-world topology involving at the same time small-scale local clusters and long-range connections between distant areas (Bassett and Bullmore, 2006). Measures at the intersection between neuroscience and complexity theory have emerged such as topological data analysis (Phinyomark et al, 2017) or multivariate auto-regressive models (Astolfi et al, 2007)

Methods

Results

Conclusion