Network Neuroscience最新文献

Recent machine learning techniques have improved connectome-based predictions by modeling complex dependencies between brain connectivity and cognitive traits. However, they typically require large datasets that are costly and time-consuming to collect. To address this, we propose Task-guided generative adversarial network (GAN) II, a novel data augmentation method that uses GANs to expand sample sizes in connectome-based prediction tasks. Our method incorporates a task-guided branch within the Wasserstein GAN framework, specifically designed to synthesize structural connectivity matrices and improve prediction accuracy by capturing task-relevant features. We evaluated Task-guided GAN II on the prediction of fluid intelligence using the NIMH Health Research Volunteer Dataset. Results showed that data augmentation improved prediction accuracy. To further assess whether augmentation can substitute for increasing actual collected sample sizes, we conducted additional validation using the Human Connectome Project WU-Minn S1200 dataset. Task-guided GAN II improved prediction performance with limited real data, with gains of up to twofold augmentation observed. However, excessive augmentation did not result in further improvements, suggesting that augmentation complements, but does not fully replace, real data augmentation. These results suggest that Task-guided GAN II is a promising tool for harnessing small datasets in human connectomics research, improving predictive modeling where large-scale data collection is impractical.

Interregional communication within the human brain is essential for maintaining functional integrity. A promising approach for investigating how brain regions communicate relies on the assumption that the brain operates as a complex network. In this context, positron emission tomography (PET) images have been suggested as a valuable source for understanding brain networks. However, such networks are typically assembled through direct computation without accounting for outliers, impacting the reliability of group representative networks. In this study, we used brain [¹⁸F]fluoro-2-deoxyglucose PET data from 1,227 individuals in the Alzheimer's disease (AD) continuum from the Alzheimer's Disease Neuroimaging Initiative cohort to develop a novel method for constructing stable metabolic brain networks that are resilient to spurious data points. Our multiple sampling scheme generates brain networks with greater stability compared with conventional approaches. The proposed method is robust to imbalanced datasets and requires 50% fewer subjects to achieve stability than the conventional method. We further validated the approach in an independent AD cohort (n = 114) from São Paulo, Brazil (Faculdade de Medicina da Universidade de São Paulo). This innovative method is flexible and improves the robustness of metabolic brain network analyses, supporting better insights into brain connectivity and resilience to data variability across multiple radiotracers for both health and disease.

Brain structure-function coupling has been studied in health and disease by many different researchers in recent years. Most of the studies have estimated functional connectivity matrices as correlation coefficients between different brain areas, despite well-known disadvantages compared with partial correlation connectivity matrices. Indeed, partial correlation represents a more sensible model for structural connectivity since, under a Gaussian approximation, it accounts only for direct dependencies between brain areas. Motivated by this and following previous results by different authors, we investigate structure-function coupling using partial correlation matrices of functional magnetic resonance imaging brain activity time series under various regularization (also known as noise-cleaning) algorithms. We find that, across different algorithms and conditions, partial correlation provides a higher match with structural connectivity retrieved from density-weighted imaging data than standard correlation, and this occurs at both subject and population levels. Importantly, we also show that regularization and thresholding are crucial for this match to emerge. Finally, we assess neurogenetic associations in relation to structure-function coupling, which presents promising opportunities to further advance research in the field of network neuroscience, particularly concerning brain disorders.

The dynamics of the human cortex are highly metastable, driving the spontaneous exploration of network states. This metastability depends on circuit-level edge-of-bifurcation dynamics, which emerge from firing-rate control through multiple mechanisms of excitatory-inhibitory (E-I) homeostasis. However, it is unclear how these contribute to the metastability of cortical networks. We propose that individual mechanisms of the E-I homeostasis contribute uniquely to the emergence of resting-state dynamics and test this hypothesis in a large-scale model of the human cortex. We show that empirical connectivity and dynamics can only be reproduced when accounting for multiple mechanisms of the E-I homeostasis. More specifically, while the homeostasis of excitation and inhibition enhances metastability, the regulation of intrinsic excitability ensures moderate synchrony, maximizing functional complexity. Furthermore, the modulation bifurcation modulation by the homeostasis of excitation and intrinsic excitability compensates for strong input fluctuations in connector hubs. Importantly, this only occurs in models accounting for local gamma oscillations, suggesting a relationship between E-I balance, gamma rhythms, and metastable dynamics. Altogether, our results show that cortical networks self-organize toward maximal metastability through the multifactor homeostasis of E-I balance. Therefore, the benefits of combining multiple homeostatic mechanisms transcend the circuit level, supporting the metastable dynamics of large-scale cortical networks.

Traumatic brain injury (TBI) is a major cause of disability leading to multiple sequelae in cognitive, sensory, and physical domains, including posttraumatic epilepsy. Despite extensive research, our understanding of its impact on macroscopic brain circuitry remains incomplete. We analyzed electrophysiological functional connectomes in the gamma band from an animal model of blast-induced TBI over multiple time points after injury. We revealed differences in small-world propensity and rich-club structure compared with age-matched controls, indicating functional reorganization following injury. We further investigated cost-efficiency trade-offs, propose a computationally efficient normalization procedure for quantifying the cost of spatially embedded networks that controls for connectivity strength differences, and observed dynamic changes across the injury timeline. To explore potential links between altered network topology and epileptic activity, we employed a brain-wide computational model of seizure dynamics and attribute brain reorganization to a homeostatic mechanism of activity regulation with the potential unintended consequence of driving generalized seizures. Finally, we demonstrated post-injury hyperexcitability that manifests as an increase in sound-evoked response amplitudes at the cortical level. Our work characterizes, for the first time, gamma-band functional network reorganization in a model of brain injury and proposes potential causes of these changes, thus identifying targets for future therapeutic interventions.

Studying structural brain networks has witnessed significant advancement in recent decades. Findings revealed a geometric principle, the exponential distance rule (EDR) showing that the number of neurons decreases exponentially with the length of their axons. This neuron-level information was used to build a region-level EDR network model that was able to explain various characteristics of interareal cortical networks in macaques, mice, and rats. The complete connectome of the Drosophila has recently been mapped providing information also about the network of neuropils (projectome). A recent study demonstrated the presence of the EDR in the Drosophila. In our study, we first revisit the EDR itself and precisely measure the characteristic decay rate. Next, we demonstrate that the EDR model effectively accounts for numerous binary and weighted properties of the projectome. Our study illustrates that the EDR model is a suitable null model for analyzing networks of brain regions, as it captures properties of region-level networks in very different species. The importance of the null model lies in its ability to facilitate the identification of functionally significant features not caused by inevitable geometric constraints, as we illustrate with the pronounced asymmetry of connection weights important for functional hierarchy.

Neuroendocrinology has received little attention in human neuroscience research, resulting in a dearth of knowledge surrounding potent and dynamic modulators of cognition and behavior, as well as brain structure and function. This work addresses one such phenomenon by studying functional connectomics related to ovarian hormone fluctuations throughout the adult menstrual cycle. To do so, we used fMRI and hormone assessments from two dense, longitudinal datasets to assess variations in functional connectivity with respect to endogenous and exogenous endocrine factors throughout the menstrual cycle. First, we replicated prior findings that common, group-level, and individual-specific factors have similar relative contributions to functional brain network organization. Second, we found widespread connectivity related to hormonal contraceptive (HC) use, in addition to sparser estradiol- and progesterone-related connectivity. Differential generalizability of these connectivity patterns suggests progestin-specific impacts on functional brain organization in HC users. These results provide novel insight into within-individual changes in brain organization across the menstrual cycle and the extent to which these changes are shared between individuals, illuminating understudied phenomena in reproductive health and important information for all neuroimaging studies that include participants who menstruate.

Single-subject morphological brain networks derived from cross-feature correlation of macroscopic MRI-derived morphological measures provide an important means for studying the brain connectome. However, the validity of this approach remains to be confirmed at the microscopic level. Here, we constructed morphological brain networks at the single-cell level by extending features from macroscopic morphological measures to microscopic descriptions of neuronal morphology. We demonstrated the feasibility and generalizability of the method using neurons in the somatosensory cortex of a rat, neurons over the whole brain of a mouse, and neurons in the middle temporal gyrus (MTG) of a human. We found that interneuron morphological similarity was higher for intra- than interclass connections, depended on cytoarchitectonic, chemoarchitectonic, and laminar classification of neurons (rat), differed between regions with different evolutionary timelines (mouse), and correlated with neuronal axonal projections (mouse). Furthermore, highly connected hub neurons were disproportionately from superficial layers (rat), inhibitory neurons (rat), and subcortical regions (mouse), and exhibited unique morphology. Finally, we demonstrated a more segregated, less integrated, and economic network architecture with worse resistance to targeted attacks for neurons in human MTG than neurons in a mouse's primary visual cortex. Overall, our method provides an alternative avenue to study neuronal wiring diagrams in brains.

In fMRI research, graphical models are used to uncover complex patterns of relationships between brain regions. Connectivity-based fMRI studies typically analyze nested data; raw observations, for example, BOLD responses, are nested within participants, which are nested within populations, for example, healthy controls. Often, studies ignore the nested structure and analyze participants either individually or in aggregate. This overlooks the distinction between within-participant and between-participant variance, which can lead to poor generalizability of results because group-level effects do not necessarily reflect effects for each member of the group and, at worst, risk paradoxical results where group-level effects are opposite to individual-level effects (e.g., Kievit, Frankenhuis, Waldorp, & Borsboom, 2013; Robinson, 2009; Simpson, 1951). To address these concerns, we propose a multilevel approach to model the fMRI networks, using a Gaussian graphical model at the individual level and a Curie-Weiss graphical model at the group level. Simulations show that our method outperforms individual or aggregate analysis in edge retrieval. We apply the proposed multilevel approach to resting-state fMRI data of 724 healthy participants, examining both their commonalities and individual differences. We not only recover the seven previously found resting-state networks at the group level but also observe considerable heterogeneity in the individual-level networks. Finally, we discuss the necessity of a multilevel approach, additional challenges, and possible future extensions.

Functional MRI (fMRI) and diffusion-weighted imaging (DWI) help explore correlations between structural connectivity (SC) and functional connectivity (FC; SC-FC coupling). Studies on mild cognitive impairment (MCI) and Alzheimer's disease (AD) observed coupling disruptions, co-occurring with cognitive decline. Advanced "fixel-based" analyses improved DWI's accuracy in assessing microstructural and macrostructural features of white matter (WM), but previous aging coupling studies commonly defined SC via tensor-based tractography and streamline counts, thereby missing fiber-specific information. We investigated different types of fixel-FC coupling and their relation to cognition in 392 participants (Age_mean = 73; 207 females) from the ADNI. Two hundred twenty-five controls, 142 MCI, and 25 AD with diffusion-weighted and resting-state fMRI scans were analyzed. Structural connectomes were constructed using average fixel metrics (fiber density (FD), fiber-bundle cross-section log, and combined [FDC]) as edges. SC-FC coupling for each SC metric was calculated at overall network, edge, and node levels. Overall DMN, node- and edge-specific coupling differences were found across SC measures and groups. DMN nodal coupling significantly predicted Mini-Mental Status Examination score and verbal memory. In conclusion, different types of fixel-based coupling alterations can be observed across the neurocognitive aging spectrum, in particular, FD-FC and FDC-FC coupling between DMN regions are associated with cognitive functioning.