Biology subjects

Luppi, A.

Publications and source records attributed to Luppi, A..

15 recordsLinked to original sources

bioRxiv2026

Psychedelics Restructure the Global Workspace of Consciousness

Psychoactive drugs profoundly alter consciousness, but the neural mechanisms by which they do so remain poorly understood. Global workspace theory may provide a framework for investigating these mechanisms. The global workspace, a candidate neural substrate of consciousness, is a set of brain regions that integrate and disseminate information to the rest of the brain. According to a recent model, the integrators, or 'gateways,' are characterized by synergistic interactions with other brain regions, while the disseminators, or 'broadcasters,' exhibit redundant interactions. Here, we measured the effects of various psychoactive drugs, including 'classic' psychedelics like LSD, mescaline, and psilocybin, 'atypical' psychedelics (entactogens) like MDMA, and the stimulant D-amphetamine, on the global workspace. In a secondary analysis of resting-state functional magnetic resonance imaging (fMRI) data on all these drugs, we found that classic psychedelics are associated with significant increases in the number of gateways in the frontoparietal control and dorsal attention networks. Meanwhile, MDMA and amphetamine reduced the number of gateways in the visual and somatomotor networks. The regional topographies of both synergy and redundancy distinguished classic psychedelics from other drugs. Finally, drugs producing similar changes in synergy and redundancy also produced similar profiles of subjective effects. Overall, our findings provide the first direct evidence that psychedelics may alter the synergistic and redundant interactions that characterize the global workspace of consciousness.

neuroscience↗

bioRxiv2026

How Similar Are Two Brains? A Comprehensive Benchmark of Brain Network Similarity Measures

A central goal of neuroscience is to establish how similar (or different) brain networks are across individuals, development, psychiatric conditions, and even between biological and artificial brains. Whenever one such comparison is made, it relies on the assumption that we have a measure that produces an accurate and plausible similarity score. Yet, the field lacks one such measure, and the wide variety of existing measures often produces conflicting results. To address this problem, we perform a systematic comparison of 16 established similarity measures. First, we show that different measures consistently disagree on which brain networks are most similar. Second, we rank all measures according to multiple criteria such as biological plausibility, computational efficiency, and sensitivity. We show that DeltaCon is the most accurate measure in terms of parameter recovery, and that it ranks among the fastest and most noise-tolerant. For this reason, we conclude that DeltaCon is the best general-purpose similarity measure. However, no measure ranks best on more than two of the five criteria we assessed, and different measures should be preferred depending on the exact research question. Measure selection can now be made on evidence and should be a crucial decision in any study that compares brain networks.

neuroscience↗

bioRxiv2026

High-order brain interactions distinguish wakefulness, anaesthesia, and recovery induced by deep brain stimulation

Understanding how consciousness depends on large-scale brain interactions is key for both the neuroscience of consciousness and clinical translation. However, it requires moving beyond classical pairwise descriptions of functional connectivity, which cannot capture the collective dependencies emerging across multiple brain regions. Here, we use multivariate information theory measures to characterize how higher-order interactions re-organize across states of consciousness. Specifically, we apply O-information to resting-state fMRI data from non-human primates to quantify whether multiregional brain dynamics are dominated by synergistic or redundant information sharing. We analyse two complementary datasets: (i) wakefulness and anaesthesia-induced loss of consciousness using different molecular agents (propofol, sevoflurane, ketamine), and (ii) the recovery of consciousness driven by central thalamic deep brain stimulation during propofol anaesthesia, indexed by behavioural responsiveness. We identify optimal regional subsets whose O-information robustly discriminates conscious from non-responsive states under two complementary optimization polarities. The first captures elevated redundancy in conscious scans that decreases under anaesthesia, providing robust discrimination and placing high-voltage central-thalamus stimulation closer to wakefulness. The second captures a synergy-to-redundancy transition, prominent in multi-anaesthesia conditions but context-dependent across datasets. Discrimination performance depends on interaction order: redundancy-based signatures improve with increasing subset size, whilst synergy-based signatures peak at low orders. Higher-order informational features significantly outperform pairwise functional connectivity, particularly for synergistic signatures which remain invisible to correlations. These findings demonstrate that consciousness is reflected in the reconfiguration of higher-order interaction structures, with distinct informational substrates requiring multivariate characterization beyond pair-wise connectivity.

neuroscience↗

bioRxiv2026

The impact of homeostatic inhibitory plasticity in a generative biophysical model

A main characteristic of biological systems is their capacity to dynamically adapt to environmental changes. In the brain, synaptic plasticity enables the strengthening or weakening of connections between neurons, allowing neural circuits to adapt based on experience, learning, and environmental changes. Yet, it is homeostatically regulated such that it avoids excessive proliferation of synaptic contacts. These mechanisms can be studied with large-scale models of brain activity. Here, we embed a biologically grounded inhibitory-homeostatic plasticity rule into the Dynamic Mean Field (DMF) model, creating a Homeostatic Dynamic Mean Field (HDMF) model that dynamically tunes local excitation-inhibition balance. Convergence of excitatory firing rates is reached by mapping a large range of coupling strength to parameters of inhibitory synapses. The HDMF reproduces statistical observables of brain activity as well as the original DMF, and can sustain neuromodulatory perturbations without overhead computations. The HDMF can generate unprecedented sleep-like slow-wave activity, which can also coexist with wake-like asynchronous dynamics, permitting to model dissociated states of consciousness such as parasomnias. Together, these results show that a single homeostatic rule broadens the stability and expressiveness of the DMF, providing a unified platform for studying how local adaptive processes shape the diverse global dynamics of the human brain.

neuroscience↗

bioRxiv2025

Human cortical networks trade communication efficiency for computational reliability.

Brains are often described as cost-efficient communication networks, optimally balancing the cost of long connections with the benefits of fast communication. Here, inspired by the "use it or lose it" principle, we present a novel game-theoretic model of self-organizing neural units and show that the brain is, in fact, sub-optimal in both regards: First, we demonstrate that regional competition for connectivity under propagative communication dynamics naturally gives rise to network configurations similar to those derived from the human cortex while being even more efficient and economical. Next, we use a reservoir computing framework to compare the information processing capacity of these networks against those of the brain. Although comparable in performance, the more optimal trade-off comes with a tax on computational reliability. Through synthetic lesions, we show that these networks are fragile because, to optimize for communication, they funnel information through a spatially clustered "oligarchy" comprising a tight set of transmodal hubs. In contrast, the human brain uses a more distributed "rich club" backbone that better resists breakdown following targeted attacks, even when it means higher wiring costs and less efficient communication. This reveals a previously overlooked principle: cortical networks trade both cost and efficiency for reliable computation. Thus, our findings highlight computational reliability as another, and even more prominent driver of brain connectivity compared to wiring cost and communication efficiency. TeaserThe brain avoids fragile efficiency: brain networks favor reliable computation, revealing resilience as a hidden design rule.

neuroscience↗

bioRxiv2025

The synergistic interactions of low-dimensional brain modes

Neuroimaging techniques produce vast amounts of data, capturing brain activity in a high-dimensional space. However, brain dynamics are consistently shown to reside in a rather lower-dimensional space, which contains relevant information for cognition and behavior. This dimensionality reduction reflects distinct types of interactions between brain regions, such as redundancy -shared neural information distributed across regions- and synergy, where information emerges only when regions are considered collectively. Significant efforts have been devoted to developing linear and non-linear algorithms to reveal these low-dimensional dynamics, often termed "brain modes." Here, we apply various dimensionality reduction techniques to resting-state functional magnetic resonance imaging (fMRI) data from 100 healthy participants to examine how synergistic interactions in brain dynamics are preserved by these techniques. We first demonstrate that biologically informed brain parcellation modulates and preserves synergy-dominated interactions. Next, we show that synergy among low-dimensional modes enhances functional-connectivity reconstruction: nonlinear autoencoders not only achieve the lowest reconstruction error but also maximally preserve synergy, outperforming principal component analysis, diffusion maps, and Laplacian eigenmodes. Finally, we confirm previous results suggesting that global signal regression helps to identify synergistic interactions between regions. Our findings establish synergy preservation as a complementary criterion to reconstruction accuracy, highlighting autoencoders as a nonlinear tool for uncovering synergistic low-dimensional brain modes from high-dimensional neuroimaging data.

neuroscience↗

bioRxiv2025

The cost of cognition: Measuring the energy consumption of non-equilibrium computation

In biological systems, survival is predicated on an animal being able to perform computations quickly on a minimal energy budget. What is the energy consumption of non-equilibrium brain computation, i.e., what is the cost of cognition? Previous literature has estimated the metabolic cost using neuroimaging measures of glucose consumption but complementary to these findings, here we directly estimate the computational costs by combining the new field of stochastic thermodynamics with whole-brain modelling. We developed the COCO (COst of COgnition) framework using an analytical expression quantifying the links between energy cost, non-equilibrium and information processing for any given brain state measured with neuroimaging. Importantly, this key relationship also holds at the level of individual brain regions. We used this to quantify the benefits of information processing on the highly anatomically, interconnected hierarchical systems of the brain. Crucially, in empirical neuroimaging data we demonstrate that the human brain uses significantly less energy overall than other mammals (including non-human primates and mice), suggestive of an evolutionary optimisation of the effectiveness of computation. Focusing on the cost of cognition, using large-scale human neuroimaging data of 970 healthy human participants, we show that the resting state uses significantly less energy that seven different cognitive tasks. Furthermore, different kinds of tasks require different amounts of non-equilibrium, information processing and energy consumption. We found that tasks requiring more distributed computation also use more energy. Overall, these results directly quantify the cost of cognition, i.e., the non-equilibrium and energetic demands of information processing, allowing a deeper understanding of how the brain compute in a way that is far more energy efficient than current generations of digital computers and artificial intelligence.

neuroscience↗

bioRxiv2025

Evolution's boldest trick: Neurotransmission modulated whole-brain computation captures full task repertoire

The perhaps most important unsolved problem in neuroscience is how the brain survives in a complex world by performing a rich repertoire of computation on a minimal energy budget. The brain is much better at adapting to the multiplicity of stimuli and outcomes than current generations of computers, artificial neural deep learning and reservoir model architectures. Yet, at first glance the brain appears to use a fixed anatomical architecture to perform the necessary huge variety of computations. But evolutions boldest trick is that in fact the brains effective connectivity is constantly being updated through neuromodulation to allow the rich repertoire of computation. Inspired by this, we created a whole-brain model using empirical neurotransmitter maps modulating the underlying local regional dynamics. This NEMO (neurotransmission modulated) whole-brain model is able to flexibly compute the full task repertoire and associated functional connectivity of the neuroimaging data from 971 healthy participants. For each individual we defined a measure of brain computability as the fitting of the NEMO whole-brain model to all tasks performed by the individual. Importantly, brain computability correlates with both behavioural performance on individual tasks and with a general behavioural measure of intelligence. Overall, our proposed unifying NEMO framework offers a natural way to sculpt different brain dynamics in a fixed brain architecture to compute the rich repertoire of tasks required for surviving and thriving.

neuroscience↗

bioRxiv2024

Time-resolved coupling between connectome harmonics and subjective experience under the psychedelic DMT

Exploring the intricate relationship between brains structure and function, and how this affects subjective experience is a fundamental pursuit in neuroscience. Psychedelic substances offer a unique insight into the influences of specific neurotransmitter systems on perception, cognition and consciousness. Specifically, their impact on brain function propagates across the structural connectome -- a network of white matter pathways linking different regions. To comprehensively grasp the effects of psychedelic compounds on brain function, we used a theoretically rigorous framework known as connectome harmonic decomposition. This framework provides a robust method to characterize how brain function intricately depends on the organized network structure of the human connectome. We show that the connectome harmonic repertoire under DMT is reshaped in line with other reported psychedelic compounds - psilocybin, LSD and ketamine. Furthermore, we show that the repertoire entropy of connectome harmonics increases under DMT, as with those other psychedelics. Importantly, we demonstrate for the first time that measures of energy spectrum difference and repertoire entropy of connectome harmonics indexes the intensity of subjective experience of the participants in a time-resolved manner reflecting close coupling between connectome harmonics and subjective experience.

neuroscience↗

bioRxiv2024

LSD flattens the hierarchy of directed information flow in fast whole-brain dynamics

Psychedelics are serotonergic drugs that profoundly alter consciousness, yet their neural mechanisms are not fully understood. A popular theory, RElaxed Beliefs Under pSychedelics (REBUS), posits that psychedelics flatten the hierarchy of information flow in the brain. Here, we investigate hierarchy based on the imbalance between sending and receiving brain signals, as determined by directed functional connectivity. We measure directed functional hierarchy in a magnetoencephalography (MEG) dataset of 16 healthy human participants who were administered a psychedelic dose (75 micrograms, intravenous) of lysergic acid diethylamide (LSD) under four different conditions. LSD diminishes the asymmetry of directed connectivity when averaged across time. Additionally, we demonstrate that machine learning classifiers distinguish between LSD and placebo more accurately when trained on one of our hierarchy metrics than when trained on traditional measures of functional connectivity. Taken together, these results indicate that LSD weakens the hierarchy of directed connectivity in the brain by increasing the balance between senders and receivers of neural signals.

neuroscience↗

bioRxiv2024

Controlling the human connectome with spatially diffuse input signals

The human brain is never at "rest"; its activity is constantly fluctuating over time, transitioning from one brain state-a whole-brain pattern of activity-to another. Network control theory offers a framework for understanding the effort - energy - associated with these transitions. One branch of control theory that is especially useful in this context is "optimal control", in which input signals are used to selectively drive the brain into a target state. Typically, these inputs are introduced independently to the nodes of the network (each input signal is associated with exactly one node). Though convenient, this input strategy ignores the continuity of cerebral cortex - geometrically, each region is connected to its spatial neighbors, allowing control signals, both exogenous and endogenous, to spread from their foci to nearby regions. Additionally, the spatial specificity of brain stimulation techniques is limited, such that the effects of a perturbation are measurable in tissue surrounding the stimulation site. Here, we adapt the network control model so that input signals have a spatial extent that decays exponentially from the input site. We show that this more realistic strategy takes advantage of spatial dependencies in structural connectivity and activity to reduce the energy (effort) associated with brain state transitions. We further leverage these dependencies to explore near-optimal control strategies such that, on a per-transition basis, the number of input signals required for a given control task is reduced, in some cases by two orders of magnitude. This approximation yields network-wide maps of input site density, which we compare to an existing database of functional, metabolic, genetic, and neurochemical maps, finding a close correspondence. Ultimately, not only do we propose a more efficient framework that is also more adherent to well-established brain organizational principles, but we also posit neurobiologically grounded bases for optimal control.

neuroscience↗

bioRxiv2023

Dynamical structure-function correlations provide robust and generalizable signatures of consciousness in humans

Resting-state functional magnetic resonance imaging evolves through a repertoire of functional connectivity patterns which might reflect ongoing cognition, as well as the contents of conscious awareness. We investigated whether the dynamic exploration of these states can provide robust and generalizable markers for the state of consciousness in human participants, across loss of consciousness induced by general anaesthesia or slow wave sleep. By clustering transient states of functional connectivity, we demonstrated that brain activity during unconsciousness is dominated by a recurrent pattern primarily mediated by structural connectivity and with a reduced capacity to transition to other patterns. Our results provide evidence supporting the pronounced differences between conscious and unconscious brain states in terms of whole-brain dynamics; in particular, the maintenance of rich brain dynamics measured by entropy is a critical aspect of conscious awareness. Collectively, our results may have significant implications for our understanding of consciousness and the neural basis of human awareness, as well as for the discovery of robust signatures of consciousness that are generalizable among different brain conditions.

neuroscience↗

bioRxiv2022

Metastability, fractal scaling, and synergistic information processing: what phase relationships reveal about intrinsic brain activity

Dynamic functional connectivity (dFC) in resting-state fMRI holds promise to deliver candidate biomarkers for clinical applications. However, the reliability and interpretability of dFC metrics remain contested. Despite a myriad of methodologies and resulting measures, few studies have combined metrics derived from different conceptualizations of brain functioning within the same analysis - perhaps missing an opportunity for improved interpretability. Using a complexity-science approach, we assessed the reliability and interrelationships of a battery of phase-based dFC metrics including tools originated from dynamical systems, stochastic processes, and information dynamics approaches. Our analysis revealed novel relationships between these metrics, which allowed us to build a predictive model for integrated information using metrics from dynamical systems and information theory. Furthermore, global metastability - a metric reflecting simultaneous tendencies for coupling and decoupling - was found to be the most representative and stable metric in brain parcellations that included cerebellar regions. Additionally, spatiotemporal patterns of phase-locking were found to change in a slow, non-random, continuous manner over time. Taken together, our findings show that the majority of characteristics of resting-state fMRI dynamics reflect an interrelated dynamical- and informational-complexity profile, which is unique to each acquisition. This finding challenges the interpretation of results from cross-sectional designs for brain neuromarker discovery, suggesting that individual life-trajectories may be more informative than sample means. HighlightsO_LISpatiotemporal patterns of phase-locking tend to be time-invariant C_LIO_LIGlobal metastability is representative and stable in a cohort of heathy young adults C_LIO_LIdFC characteristics are in general unique to any fMRI acquisition C_LIO_LIDynamical- and informational-complexity are interrelated C_LIO_LIComplexity science contributes to a coherent description of brain dynamics C_LI

neuroscience↗

bioRxiv2021

Changes in dynamic transitions between integrated and segregated states underlie visual hallucinations in Parkinson's disease

BackgroundVisual hallucinations in Parkinsons disease (PD) are transient, suggesting a change in dynamic brain states. However, the causes underlying these dynamic brain changes are not known. MethodsFocusing on fundamental network properties of integration and segregation, we used rsfMRI to examine alterations in temporal dynamics in PD patients with hallucinations (n=16) compared to those without hallucinations (n=75) and a group of normal controls (n=32). We used network control theory to examine how structural connectivity guides transitions between functional states. We then studied the brain regions most involved in these state transitions, and examined corresponding neurotransmitter density profiles and receptor gene expression in health. ResultsThere were significantly altered temporal dynamics in PD with hallucinations, with an increased proportion of time spent in the Segregated state compared to non-hallucinators and controls; less between-state transitions; and increased dwell time in the Segregated state. The energy cost needed to transition from integrated-to-segregated state was lower in PD-hallucinators compared to non-hallucinators. This was primarily driven by subcortical and transmodal cortical brain regions, including the thalamus and default mode network regions. The regional energy needed to transition from integrated-to-segregated state was significantly correlated with regional neurotransmitter density and gene expression profiles for serotoninergic (including 5HT2A), GABAergic, noradrenergic and cholinergic but not dopaminergic density profiles. ConclusionsWe describe the patterns of temporal functional dynamics in PD-hallucinations, and link these with neurotransmitter systems involved in early sensory and complex visual processing. Our findings provide mechanistic insights into visual hallucinations in PD and highlighting potential therapeutic targets.

neuroscience↗

bioRxiv2019

Consciousness & Brain Functional Complexity in Propofol Anaesthesia

The brain is possibly the most complex system known to mankind, and its complexity has been called upon to explain the emergence of consciousness. However, complexity can take many forms: here, we investigate measures of algorithmic and process complexity in both the temporal and topological dimension, testing them on functional MRI data obtained from individuals undergoing various levels of sedation with the anaesthetic agent propofol, in two separate datasets. We demonstrate that the various measures are differently able to discriminate between levels of sedation, with temporal measures showing higher sensitivity. Further, we show that all measures are strongly related to a single underlying construct explaining most of the variance, as assessed by Principal Component Analysis, which we interpret as a measure of overall complexity of our data. This overall complexity was also able to discriminate between levels of sedation, supporting the hypothesis that consciousness is related to complexity - independent of how the latter is measured.

neuroscience↗