Brain connectivity fingerprinting and behavioural prediction rest on distinct functional systems of the human connectome

Abstract

The prediction of inter-individual behavioural differences from neuroimaging data is a rapidly evolving field of research focusing on individualised methods to describe human brain organisation on the single-subject level. One method that harnesses such individual signatures is functional connectome fingerprinting, which can reliably identify individuals from large study populations. However, the precise relationship between functional signatures underlying fingerprinting and behavioural prediction remains unclear. Expanding on previous reports, here we systematically investigate the link between discrimination and prediction on different levels of brain network organisation (individual connections, network interactions, topographical organisation, and connection variability). Our analysis revealed a substantial divergence between discriminatory and predictive connectivity signatures on all levels of network organisation. Across different brain parcellations, thresholds, and prediction algorithms, we find discriminatory connections in higher-order multimodal association cortices, while neural correlates of behaviour display more variable distributions. Furthermore, we find the standard deviation of connections between participants to be significantly higher in fingerprinting than in prediction, making inter-individual connection variability a possible separating marker. These results demonstrate that participant identification and behavioural prediction involve highly distinct functional systems of the human connectome. The present study thus calls into question the direct functional relevance of connectome fingerprints.

Ritualistic use of ayahuasca enhances a shared functional connectome identity with others

“…Instead, our results were contingent on the inclusion of PC3 as a predictor of interest, suggesting higher-order PC deviations capturing individual variation in FC were most relevant to the visual and auditory effects of ayahuasca. Furthermore, predictive edges were found to span primarily both higher order systems (DMN, FPN) and primary systems (VIS, SM), with the former contributing more (in number) to behavioural prediction as per prior work 12,21,37 . Given their developmentally late maturation 79 , susceptibility to individual environmental effects 80 and, dense 5-HT2A expression 81 and coordination of multisensory integration in comparison to primary systems 61 , higher-order regions may more easily account for divergent phenomena, more so than primary systems, themselves partially influenced by the temporary states of each individual during scanning 82 .…”

“…While a subset of 250 edges could maximally define a subject's fingerprint, their importance markedly dropped under ayahuasca. Disseminated across higher-order association cortices these regions are shown to encode the majority of inter-individual variance 12,37 . Importantly, it has been previously hypothesised that the appearance of a desegregated functional architecture under psychedelics stems from the impairment of these same functional subsystems 35 .…”

“…Within a hierarchical predictive processing framework, these outcomes are hypothesised to be linked to the subjective experience via the decreased confidence in priors encoded by functional hierarchies 36 . Accordingly, it may therefore be the case that stimulation of 5-HT 2A receptors could perturb behaviourally relevant brain fingerprints otherwise residing within high-order functional networks 37 . Indeed, classical psychedelics have been speculated to be potential therapeutic interventions by improving symptomatology through rebalancing aberrant brain states 36,38,39 .…”

Intermediately Synchronised Brain States optimise trade-off between Subject Identifiability and Predictive Capacity

“…It has been suggested that the use of HACF frames with enhanced subject identifiability may amplify brain-behavior associations 24,42 . On the other hand, it has also been shown that FC-based identification and prediction may constitute conflicting goals 25,26 . Therefore, here we systematically evaluated the effect of inclusion of varying levels of functional co-fluctuations on subject identifiability and predictiveness of a range of phenotypes.…”

“…Thus far, a crucial missing component in the investigation of ETS is the evaluation of individual differences at different co-fluctuation amplitudes by means of prediction of phenotypes. Previous research has shown that connectivity of brain areas that contribute most to identification accuracy do not overlap with brain areas that contribute most to prediction accuracy 25 suggesting that FC uniqueness and stability on their own do not guarantee phenotypic relevance of brain connectivity representations 26 . Furthermore, it is possible that different subsets of frames are more or less predictive of different phenotypic domains.…”