Cervello cognitivo, come un sistema complesso interagente.

The present work investigates the challenging question of how brain physiology relates to cognitive processes, in terms of complex system physics. In the first part of our study, we focus on brain dynamics during speech or music listening, and resting state. We represent the brain as a dynamical network, where nodes correspond to different regions and links to pairwise co-fluctuations of high frequency brain activity. The evolution of this network is investigated in terms of neuronal avalanches. Both the network co-fluctuations profile and the avalanches dynamics of different subjects significantly correlate when they are listening to the same speech or music naturalistic stimulus. A detailed study on the origins of this inter-subjects correlation suggests that individual brains tend to “tick collectively” when the same stimulus is presented, and that speech and music processing may rely on a distributed network of brain regions sharing a similar dynamics and not a specific set of few specialized areas. In the second part of the present work, we focus on moving beyond the pairwise interactions, developing a computational and theoretical framework to infer task-related higher-order interactions (HOIs) in cognitive brain networks, i.e. interactions between more than two brain areas, directly related to a behavioural variable. Information theoretic measures such as O-information, redundancy, and synergy are used to overcome the challenge of going from the dynamics recorded in different brain regions to a meaningful pattern of higher-order interactions between them. We developed a procedure to simulate multivariate Gaussian data with a preset pattern of synergistic and redundant HOIs. Using this procedure we simulated a system of 12 brain regions and 1 behavioural variable. With this simulated data we successfully benchmark three different metrics of task-related HOIs. These results set the bases for a new field of research focusing on HOIs in cognitive networks, providing theoretical and computational tools for the system, cognitive and clinical neuroscience communities. Our work shows how different approaches, describing the brain as a complex system from different perspective, can provide new insights in the investigation of different research questions within the field of cognitive neuroscience.