Emergence of High-Order Functional Hubs in the Human Brain

Abstract

Network theory is often based on pairwise relationships between nodes, which is not necessarily realistic for modeling complex systems. Importantly, it does not accurately capture non-pairwise interactions in the human brain, often considered one of the most complex systems. In this work, we develop a multivariate signal processing pipeline to build high-order networks from time series and apply it to resting-state functional magnetic resonance imaging (fMRI) signals to characterize high-order communication between brain regions. We also propose connectivity and signal processing rules for building uniform hypergraphs and argue that each multivariate interdependence metric could define weights in a hypergraph. As a proof of concept, we investigate 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 find that, for each choice of multivariate interdependence, the high-order hubs are compatible with distinct systems in the brain. Additionally, the high-order functional brain networks exhibit simultaneous integration and segregation patterns qualitatively observable from their high-order hubs. Our work hereby introduces a promising heuristic route for hypergraph representation of brain activity and opens up exciting avenues for further research in high-order network neuroscience and complex systems.

Higher‐order functional connectivity analysis of resting‐state functional magnetic resonance imaging data using multivariate cumulants

“…In contrast, our work investigates higher‐order interactions as measured by multivariate cumulants. Our results suggest that higher‐order connectivity is related to the DMN, which is consistent with recent developments in information theory for neuroscience (Luppi, Mediano, Rosas, Holland, Fryer, O'Brien, et al, 2022 ; Santos et al, 2023 ). Considering the similarities between multiple approaches for quantifying higher‐order interactions, exploring their relationships in future research would be valuable.…”

Modeling the Neurocognitive Dynamics of Language across the Lifespan

“…This aligns with the DECHA model (Default Executive Coupling Hypothesis of Aging), which reports that older adults with reduced deactivation in DMN regions fail to modulate control-executive regions (Spreng & Turner, 2019;Varangis et al, 2019). This also echoes recent theoretical (Grasso et al, 2021;Varley et al, 2022) and empirical studies (Luppi et al, 2022;Santos et al, 2023;Varley et al, 2023) suggesting that higher-order interactions facilitate higher-level cognitive performances (Williams et al, 2022) such as cognitive control (Varley et al, 2023), particularly during healthy aging (Gatica et al, 2021).…”

Understanding multimorbidity requires sign-disease networks and higher-order interactions, a perspective

“…The converse is not possible: if O-information infers a synergistic relationship then there must be significant synergy. This metric has been utilized in neuroscience, in fMRI signals to characterise higher-order communication between different regions of the brain, and to capture neural spiking dynamics (Stramaglia et al, 2021;Santos et al, 2023). This information-theoretic approach captures synergy by computing (multivariate) mutual information and is model free.…”