Network theory is often based on pairwise relationships between nodes, which is not necessarily realistic for modelling complex systems. Importantly, it does not accurately capture nondyadic interactions in the human brain, often considered one of the most complex systems. In this work, we develop a multivariate signal processing pipeline that allows us to build high-order networks from resting state fMRI signals. We also propose a connectivity and signal processing rules for building uniform hyper-graphs.
As a proof of concept, We investigated the most relevant three-point interactions in the human brain by searching for high-order “hubs” in a cohort of 100 individuals from the Human Connectome Project. We found that the high-order hubs are compatible with distinct systems in the brain. Our work introduces a promising heuristic route for hyper-graph representation of brain activity. It opens up exciting avenues for further research in high-order network neuroscience and criticality in complex systems with different systems in the brain.
Find here a short video and the associated slides.