Network Neuroscience最新文献

Current techniques to estimate directed functional connectivity from magnetoencephalography (MEG) signals involve two sequential steps: (a) estimation of the sources and their amplitude time series from the MEG data and (b) estimation of directed interactions between the source time series. However, such a sequential approach is not optimal as it leads to spurious connectivity due to spatial leakage. Here, we present an algorithm to jointly estimate the source and connectivity parameters using Bayesian filtering. We refer to this new algorithm as JEDI-MEG (Joint Estimation of source Dynamics and Interactions from MEG data). By formulating a state-space model for the locations and amplitudes of a given number of sources, we show that estimation of their connections can be reduced to a system identification problem. Using simulated MEG data, we show that the joint approach provides a more accurate reconstruction of connectivity parameters than the conventional two-step approach. Using real MEG responses to visually presented faces in 16 subjects, we also demonstrate that our method gives source and connectivity estimates that are both physiologically plausible and largely consistent across subjects. In conclusion, the proposed joint estimation approach outperforms the traditional two-step approach in determining functional connectivity in MEG data.

Amyotrophic lateral sclerosis (ALS) with only motor impairment (ALS-pure motor) and the behavioral variant of frontotemporal dementia (bvFTD) are hypothesized to represent extreme ends of a disease spectrum, which encompasses ALS with cognitive/behavioral impairment (ALSci/bi). In this longitudinal magnetoencephalography (MEG) study, we investigated changes in brain network topology of ALSci/bi over time as compared with ALS-pure motor and bvFTD patients. Resting-state MEG was recorded in ALS-pure motor (n = 9), ALSci/bi (n = 16), and bvFTD (n = 16) at baseline and 5-month follow-up, projected to source space. The corrected version of the amplitude envelope correlation was applied to compute frequency-band-specific functional connectivity between brain regions, from which the backbone of the functional networks was constructed using the minimum spanning tree (MST) approach. Reference MSTs were computed based on the functional connectivity matrices for ALS-pure motor and bvFTD, against which the networks of ALSci/bi were compared. We showed that, at baseline, networks in the theta band of ALSci/bi patients were more similar to ALS-pure motor than bvFTD. At follow-up, ALSci/bi patients' beta-band network similarity had moved away from ALS-pure motor and resembled bvFTD. In conclusion, our findings suggest that brain networks of ALSci/bi patients move along the ALS-bvFTD spectrum over time, from ALS-pure motor to bvFTD-like topology.

Epilepsy affects over 50 million people worldwide, with approximately 30% experiencing drug-resistant forms that may require surgical intervention. Accurate localisation of the epileptogenic zone (EZ) is crucial for effective treatment, but how best to use intracranial EEG data to delineate the EZ remains unclear. Previous studies have used the directionality of neural activities across the brain to investigate seizure dynamics and localise the EZ. However, the different connectivity measures used across studies have often provided inconsistent insights about the direction and the localisation power of signal flow as a biomarker for EZ localisation. In a data-driven approach, this study employs a large set of 13 distinct directed connectivity measures to evaluate neural activity flow in and out the seizure onset zone (SOZ) during interictal and ictal periods. These measures test the hypotheses of "sink SOZ" (SOZ dominantly receiving neural activities during interictal periods) and "source SOZ" (SOZ dominantly transmitting activities during ictal periods). While the results were different across connectivity measures, several measures consistently supported higher connectivity directed towards the SOZ in interictal periods and higher connectivity directed away during ictal periods. Comparing six distinct metrics of node behaviour in the network, we found that SOZ separates itself from the rest of the network, allowing for the metric of "eccentricity" to localise the SOZ more accurately than any other metrics including "in strength" and "out strength." This introduced a novel biomarker for localising the SOZ, leveraging the discriminative power of directed connectivity measures in an explainable machine learning pipeline. By using a comprehensive, objective, and data-driven approach, this study addresses previously unresolved questions on the direction of neural activities in seizure organisation and sheds light on dynamics of interictal and ictal activity in focal epilepsy.

The intricate link between brain functional connectivity (FC) and structural connectivity (SC) is explored through models performing diffusion on SC to derive FC, using varied methodologies from single to multiple graph diffusion kernels. However, existing studies have not correlated diffusion scales with specific brain regions of interest (RoIs), limiting the applicability of graph diffusion. We propose a novel approach using graph diffusion wavelets to learn the appropriate diffusion scale for each RoI to accurately estimate the SC-FC mapping. Using the open Human Connectome Project dataset, we achieve an average Pearson's correlation value of 0.833, surpassing the state-of-the-art methods for the prediction of FC. It is important to note that the proposed architecture is entirely linear, computationally efficient, and notably demonstrates the power-law distribution of diffusion scales. Our results show that the bilateral frontal pole, by virtue of it having large diffusion scale, forms a large community structure. The finding is in line with the current literature on the role of the frontal pole in resting-state networks. Overall, the results underscore the potential of graph diffusion wavelet framework for understanding how the brain structure leads to FC.

The brain operates through networks of interconnected regions, which can be disrupted by glial tumors and their treatment. This study investigates associations between this altered functional network topology and cognition in gliomas. We studied 50 adult glioma survivors (>1-year posttherapy) and 50 healthy controls. Participants underwent cognitive assessments across six domains and an 8-min resting-state functional MRI. Based on the BOLD signal, partial correlations were computed among 78 brain regions. From their absolute values, whole-brain and nodal graph metrics were derived and normalized to random graphs. Group differences in whole-brain and nodal graph metrics were assessed with Mann-Whitney U tests and mixed-design analyses of variance, respectively. Metrics exhibiting significant intergroup differences were correlated with cognitive scores, with p _bonf < 0.050 indicating significance. Among controls, 8 of 78 nodes were identified as hubs. Patients exhibited significantly higher whole-brain clustering, correlating with intelligence (r(98) = -0.409, p _bonf < 0.001) and executive functioning (r(98) = 0.300, p _bonf = 0.014). Lower centrality, higher nodal clustering, and assortativity were also observed in patients, particularly in hubs, correlating with language and executive functioning, respectively (all r(98) > 0.300, p _bonf < 0.050). Glioma patients commonly experience cognitive deficits alongside posttreatment alterations in functional network topology. Alterations in clustering, assortativity, and centrality may specifically act as compensatory mechanisms, significantly influencing cognitive functioning.

Recent studies show that functional resting-state dynamics may be modeled by lattice models near criticality, such as the 2D Ising model. The Ising temperature, which is the control parameter dictating the phase transitions of the model, can provide insight into the large-scale dynamics and is being used to better understand different brain states and neurodevelopment. This period is categorized by intricate changes in the microcircuits to consolidate networks. These changes influence the macroscopic brain dynamics and also its functional relations, which can be observed in functional magnetic resonance imaging (fMRI). Therefore, this work investigates neurodevelopment through a novel method to estimate the Ising temperature of the brain from fMRI data using functional connectivity and graph neural networks trained on Ising model networks. The main finding indicates a statistically significant negative correlation between age and temperature for typically developing children (r = -0.48, p < 0.0001) and also children with attention-deficit/hyperactivity disorder (r = -0.49, p < 0.0001). This study suggests that the brain gets distant from criticality as age increases, leading to a more ordered state.

We show that sensorimotor behavior can be reliably predicted from single-trial EEG oscillations fluctuating in a coordinated manner across brain regions, frequency bands, and movement time epochs. We define high-dimensional oscillatory portraits to capture the interdependence between basic oscillatory elements, quantifying oscillations occurring in single trials at specific frequencies, locations, and time epochs. We find that the general structure of the element interdependence networks (effective connectivity) remains stable across task conditions, reflecting an intrinsic coordination architecture and responds to changes in task constraints by subtle but consistently distinct topological reorganizations. Trial categories are reliably and significantly better separated using oscillatory portraits than from the information contained in individual oscillatory elements, suggesting an interelement coordination-based encoding. Furthermore, single-trial oscillatory portrait fluctuations are predictive of fine trial-to-trial variations in movement kinematics. Remarkably, movement accuracy appears to be reflected in the capacity of the oscillatory coordination architecture to flexibly update as an effect of movement-error integration.

The discovery of resting-state networks shifted the focus from the role of local regions in cognitive tasks to the ongoing spontaneous dynamics in global networks. Recently, efforts have been invested to reduce the complexity of brain activity recordings through the application of nonlinear dimensionality reduction algorithms. Here, we investigate how the interaction between these networks emerges as an organising principle in human cognition. We combine deep variational autoencoders with computational modelling to construct a dynamical model of brain networks fitted to the whole-brain dynamics measured with functional magnetic resonance imaging (fMRI). Crucially, this allows us to infer the interaction between these networks in resting state and seven different cognitive tasks by determining the effective functional connectivity between networks. We found a high flexible reconfiguration of task-driven network interaction patterns and we demonstrate that this reconfiguration can be used to classify different cognitive tasks. Importantly, compared with using all the nodes in a parcellation, we obtain better results by modelling the dynamics of interacting networks in both model and classification performance. These findings show the key causal role of manifolds as a fundamental organising principle of brain function, providing evidence that interacting networks are the computational engines' brain during cognitive tasks.

Neural measures that predict cognitive performance are informative about the mechanisms underlying cognitive phenomena, with diagnostic potential for neuropathologies with cognitive symptoms. Among such markers, the modularity (subnetwork composition) of whole-brain functional networks is especially promising due to its longstanding theoretical foundations and recent success in predicting clinical outcomes. We used functional magnetic resonance imaging to identify whole-brain modules at rest, calculating metrics of their spatiotemporal dynamics before and after a sensorimotor learning task on which fast learning is widely believed to be supported by a cognitive strategy. We found that participants' learning performance was predicted by the degree of coordination of modular reconfiguration and the strength of recruitment and integration of networks derived during the task itself. Our findings identify these whole-brain metrics as promising network-based markers of cognition, with relevance to basic neuroscience and the potential for clinical application.

We explore the fundamental principles underlying the architecture of the human brain's structural connectome through the lens of spectral analysis of Laplacian and adjacency matrices. Building on the idea that the brain balances efficient information processing with minimizing wiring costs, our goal is to understand how the metric properties of the connectome relate to the presence of an inherent scale. We demonstrate that a simple generative model combining nonlinear preferential attachment with an exponential penalty for spatial distance between nodes can effectively reproduce several key features of the human connectome. These include spectral density, edge length distribution, eigenmode localization, local clustering, and topological properties. Additionally, we examine the finer spectral characteristics of human structural connectomes by evaluating the inverse participation ratios (IPR _q ) across various parts of the spectrum. Our analysis shows that the level statistics in the soft cluster region of the Laplacian spectrum (where eigenvalues are small) deviate from a purely Poisson distribution due to interactions between clusters. Furthermore, we identify localized modes with large IPR values in the continuous spectrum. Multiple fractal eigenmodes are found across different parts of the spectrum, and we evaluate their fractal dimensions. We also find a power-law behavior in the return probability-a hallmark of critical behavior-and conclude by discussing how our findings are related to previous conjectures that the brain operates in an extended critical phase that supports multifractality.