Functional connectivity is linked to symbolic BOLD patterns: Replication, extension, and clinical application of the human “complexome”

Functional connectivity (FC) quantifies the temporal coherence of blood-oxygen-level-dependent (BOLD) signals across brain regions. Recently, the information-theoretic “complexome” framework has linked FC to coinciding “complexity drops”: transient moments in which regional BOLD signals simultaneously become regular. Here, we replicate this relationship in an independent dataset and extend the framework by (a) integrating it with signal cofluctuation analysis; (b) extending the previous binary concept of simultaneous complexity drops to a continuous, threshold-free calculation based on a temporal unwrapping procedure; (c) providing evidence of clinical relevance in the model disease of anti-N-methyl-D-aspartate-receptor encephalitis; and (d) deriving a new measure of pairwise dissimilarity in local BOLD patterns. This “index of pattern incongruency” (IPI) explains clinically relevant FC reductions and maps onto distinct associations with cognition beyond FC. These findings show that global FC is closely related to local patterns within underlying BOLD signals, strengthening the link between complexity dynamics and the brain’s functional organization as a large-scale network. Author Summary: Understanding how local neural dynamics give rise to large-scale functional connectivity (FC) remains a central challenge in systems neuroscience. Integrating recent node- and edge-centric approaches, we present a unified model linking global FC to local neural complexity and specifically to symbolic patterns in the underlying fMRI signals. We show that FC is linked to moments of simultaneous complexity decreases, when brain signals predominantly exhibit congruent symbolic patterns. Furthermore, well-established clinical FC alterations are explained by disrupted local complexity dynamics, illustrated in anti-N-methyl-D-aspartate-receptor encephalitis as a model disease of network dysfunction. This framework quantifies the contribution of local pattern variability to global FC, providing opportunities to study network dynamics in health and disease.