Unconsciousness reconfigures modular brain network dynamics

TL;DR

This study investigates how unconscious states like deep sleep and anesthesia alter brain network dynamics, showing reduced modular integration and a smaller dynamic core, supporting the dynamic core hypothesis of consciousness.

Contribution

It introduces a new benchmark for module detection in temporal networks and applies multilayer modularity maximization to fMRI data during unconscious states.

Findings

Unconsciousness reduces network flexibility.

The size of the dynamic core decreases during unconscious states.

Results support the dynamic core hypothesis of consciousness.

Abstract

The dynamic core hypothesis posits that consciousness is correlated with simultaneously integrated and differentiated assemblies of transiently synchronized brain regions. We represented time-dependent functional interactions using dynamic brain networks, and assessed the integrityof the dynamic core by means of the flexibility and largest multilayer module of these networks. As a first step, we constrained parameter selection using a newly developed benchmark for module detection in heterogeneous temporal networks. Next, we applied a multilayer modularity maximization algorithm to dynamic brain networks computed from functional magnetic resonance imaging (fMRI) data acquired during deep sleep and under propofol anesthesia. We found that unconsciousness reconfigured network flexibility and reduced the size of the largest spatiotemporal module, which we identified with the dynamic core. Our results present a first characterization of modular brain network dynamics during states of unconsciousness measured with fMRI, adding support to the dynamic core hypothesis of human consciousness.