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 (i) integrating it with signal cofluctuation analysis through edge-timeseries, (ii) extending the previous binary concept of simultaneous complexity drops to a continuous, threshold-free calculation, (iii) providing evidence of clinical relevance in the model disease of anti-N-methyl-D-aspartate-receptor encephalitis, and (iv) deriving a novel measure of pairwise dissimilarity in local BOLD patterns. This \'index of pattern incongruency\' (IPI) explains clinically relevant FC reductions and maps onto novel associations with cognition. 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.