Abstract
Applying tools available in network science and graph theory to study brain networks has opened a new era in understanding brain mechanisms. Brain functional networks extracted from EEG time series have been frequently studied in health and diseases. In this manuscript, we studied failure resiliency of EEG-based brain functional networks. The network structures were extracted by analysing EEG time series obtained from 30 healthy subjects in resting state eyes-closed conditions. As the network structure was extracted, we measured a number of metrics related to their resiliency. In general, the brain networks showed worse resilient behaviour as compared to corresponding random networks with the same degree sequences. Brain networks had higher vulnerability than the random ones (P < 0.05), indicating that their global efficiency (i.e., communicability between the regions) is more affected by removing the important nodes. Furthermore, the breakdown happened as a result of cascaded failures in brain networks was severer (i.e., less nodes survived) as compared to randomized versions (P < 0.05). These results suggest that real EEG-based networks have not been evolved to possess optimal resiliency against failures.
Highlights
Networked structures are abundant and many real-world systems can be modelled as networks with nodes representing the individual units and edges representing the relations between them
Our results showed that EEG-based brain functional networks are less efficient than random networks
Our results indicate that brain networks have not been evolved in a way to be optimal for communicability of its regions with each other, or at least, there are other mechanisms controlling the evolution of brain networks
Summary
Networked structures are abundant and many real-world systems can be modelled as networks with nodes representing the individual units and edges representing the relations between them. With tremendous progress in computing tools and database systems, techniques developed in Network Science have been applied to many real-world systems [1,2,3]. The brain can be described as a networked structure at both micro and macro levels. Nodes represent the defined brain regions and the edges correspond to anatomical/ functional relations between these regions. Anatomical brain networks can be studied using diffusion tensor imaging (DTI) [6] while techniques such as functional magnetic resonance imaging (fMRI), electroencephalography (EEG) and magnetocephalography (MEG) can be used to discover functional brain networks [7,8,9]. Analysis of brain networks in health and disease has revealed that their structure might be disrupted in brain disorders such as epilepsy
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
