A fundamental question in neuroscience is how brain organisation gives rise to humans' unique cognitive abilities. Although complex cognition is widely assumed to rely on frontal and parietal brain regions, the underlying mechanisms remain elusive: current approaches are unable to disentangle different forms of information processing in the brain. Here, we introduce a powerful framework to identify synergistic and redundant contributions to neural information processing and cognition. Leveraging multimodal data including functional MRI, PET, cytoarchitectonics and genetics, we reveal that synergistic interactions are the fundamental drivers of complex human cognition. Whereas redundant information dominates sensorimotor areas, synergistic activity is closely associated with the brain's prefrontal-parietal and default networks; furthermore, meta-analytic results demonstrate a close relationship between high-level cognitive tasks and synergistic information. From an evolutionary perspective, the human brain exhibits higher prevalence of synergistic information than non-human primates. At the macroscale, we demonstrate that high-synergy regions underwent the highest degree of evolutionary cortical expansion. At the microscale, human-accelerated genes promote synergistic interactions by enhancing synaptic transmission. These convergent results provide critical insights that synergistic neural interactions underlie the evolution and functioning of humans' sophisticated cognitive abilities, and demonstrate the power of our widely applicable information decomposition framework.

A central goal of neuroscience is to understand how the brain synthesises information from multiple inputs to give rise to a unified conscious experience. This process is widely believed to require integration of information. Here, we combine information theory and network science to address two fundamental questions: how is the human information-processing architecture functionally organised? And how does this organisation support human consciousness? To address these questions, we leverage the mathematical framework of Integrated Information Decomposition to delineate a cognitive architecture wherein specialised modules interact with a “synergistic global workspace,” comprising functionally distinct gateways and broadcasters. Gateway regions gather information from the specialised modules for processing in the synergistic workspace, whose contents are then further integrated to later be made widely available by broadcasters. Through data-driven analysis of resting-state functional MRI, we reveal that gateway regions correspond to the brain’s well-known default mode network, whereas broadcasters of information coincide with the executive control network. Demonstrating that this synergistic workspace supports human consciousness, we further apply Integrated Information Decomposition to BOLD signals to compute integrated information across the brain. By comparing changes due to propofol anaesthesia and severe brain injury, we demonstrate that most changes in integrated information happen within the synergistic workspace. Furthermore, it was found that loss of consciousness corresponds to reduced integrated information between gateway, but not broadcaster, regions of the synergistic workspace. Thus, loss of consciousness may coincide with breakdown of information integration by this synergistic workspace of the human brain. Together, these findings demonstrate that refining our understanding of information-processing in the human brain through Integrated Information Decomposition can provide powerful insights into the human neurocognitive architecture, and its role in supporting consciousness.

The human brain generates a rich repertoire of spatiotemporal dynamics during normal wakefulness, supporting a wide variety of conscious experiences and cognitive functions. However, neural dynamics are reconfigured, in comparable ways, when consciousness is lost either due to anaesthesia or disorders of consciousness (DOC). Here, leveraging a neurobiologically realistic whole-brain computational model informed by functional MRI, diffusion MRI, and PET, we sought to identify the neurobiological mechanisms that explain the common reconfiguration of neural dynamics observed both for transient pharmacological intervention and chronic neuroanatomical injury. Our results show that, by incorporating local inhibitory action through a PET-based GABA receptor density map, our model can reproduce the brain dynamics of subjects undergoing propofol anaesthesia, and that this effect depends specifically on the spatial distribution of GABA receptors across cortical regions. Additionally, using a structural connectome obtained from DOC patients, we demonstrate how the dynamics that characterise loss of consciousness can emerge from changes in neuroanatomical connectivity. Crucially, we find that each of these two interventions generalises across datasets: a model with increased GABA-mediated inhibition can reproduce the dynamics of DOC patients' brains, and a model with a DOC connectome is also compatible with brain dynamics observed during propofol anaesthesia. These results demonstrate how increased inhibition and connectome randomisation represent different neurobiological paths towards the characteristic dynamics of the unconscious brain. Overall, the present findings begin to disentangle the neurobiological mechanisms by which highly dissimilar perturbations of the brain's neurodynamics can lead to unconsciousness.

The functional interactions between regions of the human brain can be viewed as a network, empowering neuroscientists to leverage tools such as graph theory to obtain insight about brain function. However, obtaining a brain network from functional neuroimaging data inevitably involves multiple steps of data manipulation, which can affect the organisation (topology) of the resulting network and its properties. Test-retest reliability is a gold standard for both basic research and clinical use: a suitable data-processing pipeline for brain networks should recover the same network topology across repeated scan sessions of the same individual. Analyzing resting-state functional Magnetic Resonance Imaging (rs-fMRI) recordings from two test-retest studies across short (45 minutes), medium (2-4 weeks) and long term delays (5-16 months), we investigated the reliability of network topologies constructed by applying 576 unique pipelines to the same fMRI data, obtained from considering combinations of atlas type and size, edge definition and thresholding, and use of global signal regression. We adopted the portrait divergence, an information-theoretic criterion to measure differences in network topology across all scales, enabling us to quantify the influence of different pipelines on the overall organisation of the resulting network. Remarkably, our findings reveal that the choice of pipeline plays a fundamental role in determining how reproducible an individual's brain network topology will be across different scans: there is large and systematic variability across pipelines, such that an inappropriate choice of pipeline can distort the resulting network more than an interval of several months between scans. Across datasets and time-spans, we also identify specific combinations of data-processing steps that consistently yield networks with reproducible topology, enabling us to make recommendations about best practices to ensure high-quality brain networks.