Network Neuroscience最新文献

Sustained attention is essential for daily life and can be directed to information from different perceptual modalities, including audition and vision. Recently, cognitive neuroscience has aimed to identify neural predictors of behavior that generalize across datasets. Prior work has shown strong generalization of models trained to predict individual differences in sustained attention performance from patterns of fMRI functional connectivity. However, it is an open question whether predictions of sustained attention are specific to the perceptual modality in which they are trained. In the current study, we test whether connectome-based models predict performance on attention tasks performed in different modalities. We show first that a predefined network trained to predict adults' visual sustained attention performance generalizes to predict auditory sustained attention performance in three independent datasets (N ₁ = 29, N ₂ = 60, N ₃ = 17). Next, we train new network models to predict performance on visual and auditory attention tasks separately. We find that functional networks are largely modality general, with both model-unique and shared model features predicting sustained attention performance in independent datasets regardless of task modality. Results support the supposition that visual and auditory sustained attention rely on shared neural mechanisms and demonstrate robust generalizability of whole-brain functional network models of sustained attention.

Brain tumors can induce pathological changes in neuronal dynamics that are reflected in functional connectivity measures. Here, we use a whole-brain modeling approach to investigate pathological alterations to neuronal activity in glioma patients. By fitting a Hopf whole-brain model to empirical functional connectivity, we investigate glioma-induced changes in optimal model parameters. We observe considerable differences in neuronal dynamics between glioma patients and healthy controls, both on an individual and population-based level. In particular, model parameter estimation suggests that local tumor pathology causes changes in brain dynamics by increasing the influence of interregional interactions on global neuronal activity. Our approach demonstrates that whole-brain models provide valuable insights for understanding glioma-associated alterations in functional connectivity.

Amyloid-β (Aβ) plaques in conjunction with hyperphosphorylated tau proteins in the form of neurofibrillary tangles are the two neuropathological hallmarks of Alzheimer's disease. It is well-known that the identification of individuals with Aβ positivity could enable early diagnosis. In this work, we aim at capturing the Aβ positivity status in an unbalanced cohort enclosing subjects at different disease stages, exploiting the underlying structural and connectivity disease-induced modulations as revealed by structural, functional, and diffusion MRI. Of note, due to the unbalanced cohort, the outcomes may be guided by those factors rather than amyloid accumulation. The partial views provided by each modality are integrated in the model, allowing to take full advantage of their complementarity in encoding the effects of the Aβ accumulation, leading to an accuracy of 0.762 ± 0.04. The specificity of the information brought by each modality is assessed by post hoc explainability analysis (guided backpropagation), highlighting the underlying structural and functional changes. Noteworthy, well-established biomarker key regions related to Aβ deposition could be identified by all modalities, including the hippocampus, thalamus, precuneus, and cingulate gyrus, witnessing in favor of the reliability of the method as well as its potential in shedding light on modality-specific possibly unknown Aβ deposition signatures.

Despite significant research, discovering causal relationships from fMRI remains a challenge. Popular methods such as Granger causality and dynamic causal modeling fall short in handling contemporaneous effects and latent common causes. Methods from causal structure learning literature can address these limitations but often scale poorly with network size and need acyclicity. In this study, we first provide a taxonomy of existing methods and compare their accuracy and efficiency on simulated fMRI from simple topologies. This analysis demonstrates a pressing need for more accurate and scalable methods, motivating the design of Causal discovery for Large-scale Low-resolution Time-series with Feedback (CaLLTiF). CaLLTiF is a constraint-based method that uses conditional independence between contemporaneous and lagged variables to extract causal relationships. On simulated fMRI from the macaque connectome, CaLLTiF achieves significantly higher accuracy and scalability than all tested alternatives. From resting-state human fMRI, CaLLTiF learns causal connectomes that are highly consistent across individuals, show clear top-down flow of causal effect from attention and default mode to sensorimotor networks, exhibit Euclidean distance dependence in causal interactions, and are highly dominated by contemporaneous effects. Overall, this work takes a major step in enhancing causal discovery from whole-brain fMRI and defines a new standard for future investigations.

The functional and cognitive effects of traumatic brain injury (TBI) are poorly understood, as even mild injuries (concussion) can lead to long-lasting, untreatable symptoms. Simplified brain dynamics models may help researchers better understand the relationship between brain injury patterns and functional outcomes. Properly developed, these computational models provide an approach to investigate the effects of both computational and in vivo injury on simulated dynamics and cognitive function, respectively, for model organisms. In this study, we apply the Kuramoto model and an existing mesoscale mouse brain structural network to develop a simplified computational model of mouse brain dynamics. We explore how to optimize our initial model to predict existing mouse brain functional connectivity collected from mice under various anesthetic protocols. Finally, to determine how strongly the changes in our optimized models' dynamics can predict the extent of a brain injury, we investigate how our simulations respond to varying levels of structural network damage. Results predict a mixture of hypo- and hyperconnectivity after experimental TBI, similar to results in TBI survivors, and also suggest a compensatory remodeling of connections that may have an impact on functional outcomes after TBI.

Magnetoencephalography (MEG) is widely used for studying resting-state brain connectivity. However, MEG source imaging is ill posed and has limited spatial resolution. This introduces source-leakage issues, making it challenging to interpret MEG-derived connectivity in resting states. To address this, we validated MEG-derived connectivity from 45 healthy participants using a normative intracranial EEG (iEEG) atlas. The MEG inverse problem was solved using the wavelet-maximum entropy on the mean method. We computed four connectivity metrics: amplitude envelope correlation (AEC), orthogonalized AEC (OAEC), phase locking value (PLV), and weighted-phase lag index (wPLI). We compared spatial correlation between MEG and iEEG connectomes across standard canonical frequency bands. We found moderate spatial correlations between MEG and iEEG connectomes for AEC and PLV. However, when considering metrics that correct/remove zero-lag connectivity (OAEC/wPLI), the spatial correlation between MEG and iEEG connectomes decreased. MEG exhibited higher zero-lag connectivity compared with iEEG. The correlations between MEG and iEEG connectomes suggest that relevant connectivity patterns can be recovered from MEG. However, since these correlations are moderate/low, MEG connectivity results should be interpreted with caution. Metrics that correct for zero-lag connectivity show decreased correlations, highlighting a trade-off; while MEG may capture more connectivity due to source-leakage, removing zero-lag connectivity can eliminate true connections.

Synaptic connectivity at the neuronal level is characterized by highly nonrandom features. Hypotheses about their role can be developed by correlating structural metrics to functional features. But, to prove causation, manipulations of connectivity would have to be studied. However, the fine-grained scale at which nonrandom trends are expressed makes this approach challenging to pursue experimentally. Simulations of neuronal networks provide an alternative route to study arbitrarily complex manipulations in morphologically and biophysically detailed models. Here, we present Connectome-Manipulator, a Python framework for rapid connectome manipulations of large-scale network models in Scalable Open Network Architecture TemplAte (SONATA) format. In addition to creating or manipulating the connectome of a model, it provides tools to fit parameters of stochastic connectivity models against existing connectomes. This enables rapid replacement of any existing connectome with equivalent connectomes at different levels of complexity, or transplantation of connectivity features from one connectome to another, for systematic study. We employed the framework in the detailed model of the rat somatosensory cortex in two exemplary use cases: transplanting interneuron connectivity trends from electron microscopy data and creating simplified connectomes of excitatory connectivity. We ran a series of network simulations and found diverse shifts in the activity of individual neuron populations causally linked to these manipulations.

Understanding the complex neural mechanisms underlying speech and music perception remains a multifaceted challenge. In this study, we investigated neural dynamics using human intracranial recordings. Employing a novel approach based on low-dimensional reduction techniques, the Manifold Density Flow (MDF), we quantified the complexity of brain dynamics during naturalistic speech and music listening and during resting state. Our results reveal higher complexity in patterns of interdependence between different brain regions during speech and music listening compared with rest, suggesting that the cognitive demands of speech and music listening drive the brain dynamics toward states not observed during rest. Moreover, speech listening has more complexity than music, highlighting the nuanced differences in cognitive demands between these two auditory domains. Additionally, we validated the efficacy of the MDF method through experimentation on a toy model and compared its effectiveness in capturing the complexity of brain dynamics induced by cognitive tasks with another established technique in the literature. Overall, our findings provide a new method to quantify the complexity of brain activity by studying its temporal evolution on a low-dimensional manifold, suggesting insights that are invisible to traditional methodologies in the contexts of speech and music perception.

Our brains operate as a complex network of interconnected neurons. To gain a deeper understanding of this network architecture, it is essential to extract simple rules from its intricate structure. This study aimed to compress and simplify the architecture, with a particular focus on interpreting patterns of functional connectivity in 2.5 hr of electrical activity from a vast number of neurons in acutely sliced mouse brains. Here, we combined two distinct methods together: automatic compression and network analysis. Firstly, for automatic compression, we trained an artificial neural network named NNE (neural network embedding). This allowed us to reduce the connectivity to features, be represented only by 13% of the original neuron count. Secondly, to decipher the topology, we concentrated on the variability among the compressed features and compared them with 15 distinct network metrics. Specifically, we introduced new metrics that had not previously existed, termed as indirect-adjacent degree and neighbor hub ratio. Our results conclusively demonstrated that these new metrics could better explain approximately 40%-45% of the features. This finding highlighted the critical role of NNE in facilitating the development of innovative metrics, because some of the features extracted by NNE were not captured by the currently existed network metrics.

Generative models of brain activity have been instrumental in testing hypothesized mechanisms underlying brain dynamics against experimental datasets. Beyond capturing the key mechanisms underlying spontaneous brain dynamics, these models hold an exciting potential for understanding the mechanisms underlying the dynamics evoked by targeted brain stimulation techniques. This paper delves into this emerging application, using concepts from dynamical systems theory to argue that the stimulus-evoked dynamics in such experiments may be shaped by new types of mechanisms distinct from those that dominate spontaneous dynamics. We review and discuss (a) the targeted experimental techniques across spatial scales that can both perturb the brain to novel states and resolve its relaxation trajectory back to spontaneous dynamics and (b) how we can understand these dynamics in terms of mechanisms using physiological, phenomenological, and data-driven models. A tight integration of targeted stimulation experiments with generative quantitative modeling provides an important opportunity to uncover novel mechanisms of brain dynamics that are difficult to detect in spontaneous settings.