Dynamical Complexity in the C.elegans Neural Network
arXiv Preprint Archive October 25, 2015 Chris G. Antonopoulos, Athanasios S. Fokas, Tassos C. Bountis
A computational model of the C. elegans neuronal circuit, built from Hindmarsh-Rose differential equations and divided into six communities identified by walktrap and Louvain methods, shows that the brain network generates more information than the sum of its parts. The auto-regressive integrated information measure Φ_AR, proposed as an indicator of consciousness levels, reveals that integrated information is highest when either all communities are highly synchronized or when there is a mix of highly synchronized and desynchronized communities. Both conditions correspond to relatively low chaotic behavior.