Functional brain networks reflect spatial and temporal autocorrelation

Application of neurite orientation dispersion and density imaging in characterizing brain microstructural changes in classical trigeminal neuralgia and a comparison between the left and right sides

ABSTRACT

Diffusion tensor imaging can detect brain white matter changes in classical trigeminal neuralgia (TN). However, it lacks specificity for individual tissue microstructural features, such as neurite density, orientation dispersions, and extracellular edema. Neurite orientation dispersion and density imaging (NODDI), a novel diffusion magnetic resonance imaging (MRI) technique, can provide these distinct indices. We characterized brain microstructural alterations in patients with unilateral TN using NODDI and compared the difference between left- and right-side TN (LTN and RTN, respectively). Diffusion-weighted imaging was performed on 39 patients with LTN, 37 patients with RTN, and 37 healthy controls. Neurite orientation dispersion and density imaging-related indices, including the intracellular volume fraction (Vic), orientation dispersion index (ODI), and isotropic volume fraction (Fiso), were estimated and compared using tract-based spatial statistics and voxel-based region-of-interest analysis. The LTN and RTN groups exhibited microstructural abnormalities in white and gray matter as measured by decreased fractional anisotropy and Vic and elevated Fiso, respectively. These alterations were associated with clinical features and were mainly located in the frontal lobe, corona radiata, internal capsule, and thalamus. The angular variation of neurites, characterized by ODI, exhibited no significant changes between TN and control groups. Patients with classical TN of either side exhibited reduced Vic and increased Fiso, which indicated decreased density of axons and dendrites and neuroinflammatory edema in bilateral hemispheres. Neurite orientation dispersion and density imaging is a useful technique for in vivo diffusion MRI of white and gray matter in the brain, which provides additional metrics and information closely related to the tissue microstructure that merits further study of TN pathogenesis.

Sleep Modulates Neural Timescales and Spatiotemporal Integration in the Human Cortex

Spontaneous neural dynamics manifest across multiple temporal and spatial scales, which are thought to be intrinsic to brain areas and exhibit hierarchical organization across the cortex. In wake, a hierarchy of timescales is thought to naturally emerge from microstructural properties, gene expression, and recurrent connections. A fundamental question is timescales' organization and changes in sleep, where physiological needs are different. Here, we describe two measures of neural timescales, obtained from broadband activity and gamma power, which display complementary properties. We leveraged intracranial electroencephalography in 106 human epilepsy patients (48 females) to characterize timescale changes from wake to sleep across the cortical hierarchy. We show that both broadband and gamma timescales are globally longer in sleep than in wake. While broadband timescales increase along the sensorimotor-association axis, gamma ones decrease. During sleep, slow waves can explain the increase of broadband and gamma timescales, but only broadband ones show a positive association with slow-wave density across the cortex. Finally, we characterize spatial correlations and their relationship with timescales as a proxy for spatiotemporal integration, finding high integration at long distances in wake for broadband and at short distances in sleep for gamma timescales. Our results suggest that mesoscopic neural populations possess different timescales that are shaped by anatomy and are modulated by the sleep/wake cycle.

Temporal autocorrelation is predictive of age-An extensive MEG time-series analysis

Understanding the evolving dynamics of the brain throughout life is pivotal for anticipating and evaluating individual health. While previous research has described age effects on spectral properties of neural signals, it remains unclear which ones are most indicative of age-related processes. This study addresses this gap by analyzing resting-state data obtained from magnetoencephalography (MEG) in 350 adults (18 to 88 y). We employed advanced time-series analysis at the brain region level and machine learning to predict age. While traditional spectral features achieved low to moderate accuracy, over a hundred time-series features proved superior. Notably, temporal autocorrelation (AC) emerged as the most robust predictor of age. Distinct patterns of AC within the visual and temporal cortex were most informative, offering a versatile measure of age-related signal changes for comprehensive health assessments based on brain activity.

Representing Brain-Behavior Associations by Retaining High-Motion Minoritized Youth

BACKGROUND

Population neuroscience datasets provide an opportunity for researchers to estimate reproducible effect sizes for brain-behavior associations because of their large sample sizes. However, these datasets undergo strict quality control to mitigate sources of noise, such as head motion. This practice often excludes a disproportionate number of minoritized individuals.

METHODS

We used motion-ordering and motion-ordering+resampling (bagging) to test whether these methods preserve functional magnetic resonance imaging (fMRI) data in the Adolescent Brain Cognitive Development (ABCD) Study (N = 5733). For the 2 methods, brain-behavior associations were computed as the partial Spearman's rank correlations (Rs) between functional connectivity and cognitive performance (NIH Cognition Toolbox) as well as externalizing and internalizing psychopathology (Child Behavior Checklist) while adjusting for participant sex assigned at birth and head motion.

RESULTS

Black and Hispanic youth exhibited excess head motion relative to data collected from White youth and were discarded disproportionately when conventional approaches were used. Motion-ordering and bagging methods retained more than 99% of Black and Hispanic youth. Both methods produced reproducible brain-behavior associations across low-/high-motion racial/ethnic groups based on motion-limited fMRI data.

CONCLUSIONS

The motion-ordering and bagging methods are 2 feasible approaches that can enhance sample representation for testing brain-behavior associations and that result in reproducible effect sizes in diverse populations.

Mapping and modeling age-related changes in intrinsic neural timescales

Intrinsic timescales of brain regions exhibit heterogeneity, escalating with hierarchical levels, and are crucial for the temporal integration of external stimuli. Aging, often associated with cognitive decline, involves progressive neuronal and synaptic loss, reshaping brain structure and dynamics. However, the impact of these structural changes on temporal coding in the aging brain remains unclear. We mapped intrinsic timescales and gray matter volume (GMV) using magnetic resonance imaging (MRI) in young and elderly adults. We found shorter intrinsic timescales across multiple large-scale functional networks in the elderly cohort, and a significant positive association between intrinsic timescales and GMV. Additionally, age-related decline in performance on visual discrimination tasks was linked to a reduction in intrinsic timescales in the cuneus. To explain these age-related shifts, we developed an age-dependent spiking neuron network model. In younger subjects, brain regions were near a critical branching regime, while regions in elderly subjects had fewer neurons and synapses, pushing the dynamics toward a subcritical regime. The model accurately reproduced the empirical results, showing longer intrinsic timescales in young adults due to critical slowing down. Our findings reveal how age-related structural brain changes may drive alterations in brain dynamics, offering testable predictions and informing possible interventions targeting cognitive decline.

Anchoring functional connectivity to individual sulcal morphology yields insights in a pediatric study of reasoning

A salient neuroanatomical feature of the human brain is its pronounced cortical folding, and there is mounting evidence that sulcal morphology is relevant to functional brain architecture and cognition. However, our understanding of the relationships between sulcal anatomy, brain activity, and behavior is still in its infancy. We previously found the depth of three small, shallow sulci in lateral prefrontal cortex (LPFC) was linked to reasoning performance in childhood and adolescence (Voorhies et al., 2021). These findings beg the question: what is the linking mechanism between sulcal morphology and cognition? To shed light on this question, we investigated functional connectivity among sulci in LPFC and lateral parietal cortex (LPC). We leveraged manual parcellations (21 sulci/hemisphere, total of 1806) and functional magnetic resonance (fMRI) data from a reasoning task from 43 participants aged 7-18 years (20 female). We conducted clustering and classification analyses of individual-level functional connectivity among sulci. Broadly, we found that 1) the connectivity patterns of individual sulci could be differentiated - and more accurately than rotated sulcal labels equated for size and shape; 2) sulcal connectivity did not consistently correspond with that of probabilistic labels or large-scale networks; 3) sulci clustered together into groups with similar patterns, not dictated by spatial proximity; and 4) across individuals, greater depth was associated with higher network centrality for several sulci under investigation. These results highlight that functional connectivity can be meaningfully anchored to individual sulcal anatomy, and demonstrate that functional network centrality can vary as a function of sulcal depth.

Heritability of movie-evoked brain activity and connectivity

The neural bases of sensory processing are conserved across people, but no two individuals experience the same stimulus in exactly the same way. Recent work has established that the idiosyncratic nature of subjective experience is underpinned by individual variability in brain responses to sensory information. However, the fundamental origins of this individual variability have yet to be systematically investigated. Here, we establish a genetic basis for individual differences in sensory processing by quantifying (1) the heritability of high-dimensional brain responses to movies and (2) the extent to which this heritability is grounded in lower-level aspects of brain function. Specifically, we leverage 7T fMRI data collected from a twin sample to first show that movie-evoked brain activity and connectivity patterns are heritable across the cortex. Next, we use hyperalignment to decompose this heritability into genetic similarity in where vs. how sensory information is processed. Finally, we show that the heritability of brain activity patterns can be partially explained by the heritability of the neural timescale, a one-dimensional measure of local circuit functioning. These results demonstrate that brain responses to complex stimuli are heritable, and that this heritability is due, in part, to genetic control over stable aspects of brain function.

The application value of Rs-fMRI-based machine learning models for differentiating mild cognitive impairment from Alzheimer's disease: a systematic review and meta-analysis

BACKGROUND

Various machine learning (ML) models based on resting-state functional MRI (Rs-fMRI) have been developed to facilitate differential diagnosis of mild cognitive impairment (MCI) and Alzheimer's disease (AD). However, the diagnostic accuracy of such models remains understudied. Therefore, we conducted this systematic review and meta-analysis to explore the diagnostic accuracy of Rs-fMRI-based radiomics in differentiating MCI from AD.

METHODS

PubMed, Embase, Cochrane, and Web of Science were searched from inception up to February 8, 2024, to identify relevant studies. Meta-analysis was conducted using a bivariate mixed-effects model, and sub-group analyses were carried out by the types of ML tasks (binary classification and multi-class classification tasks).

FINDINGS

In total, 23 studies, comprising 5,554 participants were enrolled in the study. In the binary classification tasks (twenty studies), the diagnostic accuracy of the ML model for AD was 0.99 (95%CI: 0.34 ~ 1.00), with a sensitivity of 0.94 (95%CI: 0.89 ~ 0.97) and a specificity of 0.98 (95%CI: 0.95 ~ 1.00). In the multi-class classification tasks (six studies), the diagnostic accuracy of the ML model was 0.98 (95%CI: 0.98 ~ 0.99) for NC, 0.96 (95%CI: 0.96 ~ 0.96) for early mild cognitive impairment (EMCI), 0.97 (95%CI: 0.96 ~ 0.97) for late mild cognitive impairment (LMCI), and 0.95 (95%CI: 0.95 ~ 0.95) for AD.

CONCLUSIONS

The Rs-fMRI-based ML model can be adapted to multi-class classification tasks. Therefore, multi-center studies with large samples are needed to develop intelligent application tools to promote the development of intelligent ML models for disease diagnosis.

A Deep Dynamic Causal Learning Model to Study Changes in Dynamic Effective Connectivity During Brain Development

OBJECTIVE

Brain dynamic effective connectivity (dEC), characterizes the information transmission patterns between brain regions that change over time, which provides insight into the biological mechanism underlying brain development. However, most existing methods predominantly capture fixed or temporally invariant EC, leaving dEC largely unexplored.

METHODS

Herein we propose a deep dynamic causal learning model specifically designed to capture dEC. It includes a dynamic causal learner to detect time-varying causal relationships from spatio-temporal data, and a dynamic causal discriminator to validate these findings by comparing original and reconstructed data.

RESULTS

Our model outperforms established baselines in the accuracy of identifying dynamic causalities when tested on the simulated data. When applied to the Philadelphia Neurodevelopmental Cohort, the model uncovers distinct patterns in dEC networks across different age groups. Specifically, the evolution process of brain dEC networks in young adults is more stable than in children, and significant differences in information transfer patterns exist between them.

CONCLUSION

This study highlights the brain's developmental trajectory, where networks transition from undifferentiated to specialized structures with age, in accordance with the improvement of an individual's cognitive and information processing capability.

SIGNIFICANCE

The proposed model consists of the identification and verification of dynamic causality, utilizing the spatio-temporal fusing information from fMRI. As a result, it can accurately detect dEC and characterize its evolution over age.

A hypergraph transformer method for brain disease diagnosis

Objective

To address the high-order correlation modeling and fusion challenges between functional and structural brain networks.

Method

This paper proposes a hypergraph transformer method for modeling high-order correlations between functional and structural brain networks. By utilizing hypergraphs, we can effectively capture the high-order correlations within brain networks. The Transformer model provides robust feature extraction and integration capabilities that are capable of handling complex multimodal brain imaging.

Results

The proposed method is evaluated on the ABIDE and ADNI datasets. It outperforms all the comparison methods, including traditional and graph-based methods, in diagnosing different types of brain diseases. The experimental results demonstrate its potential and application prospects in clinical practice.

Conclusion

The proposed method provides new tools and insights for brain disease diagnosis, improving accuracy and aiding in understanding complex brain network relationships, thus laying a foundation for future brain science research.

Measuring the dynamic balance of integration and segregation underlying consciousness, anesthesia, and sleep in humans

Consciousness requires a dynamic balance of integration and segregation in brain networks. We report an fMRI-based metric, the integration-segregation difference (ISD), which captures two key network properties: network efficiency (integration) and clustering (segregation). With this metric, we quantify brain state transitions from conscious wakefulness to unresponsiveness induced by the anesthetic propofol. The observed changes in ISD suggest a profound shift towards the segregation of brain networks during anesthesia. A common unimodal-transmodal sequence of disintegration and reintegration occurs in brain networks during, respectively, loss and return of responsiveness. Machine learning models using integration and segregation data accurately identify awake vs. unresponsive states and their transitions. Metastability (dynamic recurrence of non-equilibrium transient states) is more effectively explained by integration, while complexity (diversity of neural activity) is more closely linked with segregation. A parallel analysis of sleep states produces similar findings. Our results demonstrate that the ISD reliably indexes states of consciousness.

Neural Correlates of Psychedelic, Sleep, and Sedated States Support Global Theories of Consciousness

Understanding neural mechanisms of consciousness remains a challenging question in neuroscience. A central debate in the field concerns whether consciousness arises from global interactions that involve multiple brain regions or focal neural activity, such as in sensory cortex. Additionally, global theories diverge between the Global Neuronal Workspace (GNW) hypothesis, which emphasizes frontal and parietal areas, and the Integrated Information Theory (IIT), which focuses on information integration within posterior cortical regions. To disentangle the global vs. local and frontoparietal vs. posterior dilemmas, we measured global functional connectivity and local neural synchrony with functional magnetic resonance imaging (fMRI) data across a spectrum of conscious states in humans induced by psychedelics, sleep, and deep sedation. We found that psychedelic states are associated with increased global functional connectivity and decreased local neural synchrony. In contrast, non-REM sleep and deep sedation displayed the opposite pattern, suggesting that consciousness arises from global brain network interactions rather than localized activity. This mirror-image pattern between enhanced and diminished states was observed in both anterior-posterior (A-P) and posterior-posterior (P-P) brain regions but not within the anterior part of the brain alone. Moreover, anterior transmodal regions played a key role in A-P connectivity, while both posterior transmodal and posterior unimodal regions were critical for P-P connectivity. Overall, these findings provide empirical evidence supporting global theories of consciousness in relation to varying states of consciousness. They also bridge the gap between two prominent theories, GNW and IIT, by demonstrating how different theories can converge on shared neuronal mechanisms.

Multiple Intrinsic Timescales Govern Distinct Brain States in Human Sleep

Human sleep exhibits multiple, recurrent temporal regularities, ranging from circadian rhythms to sleep stage cycles and neuronal oscillations during nonrapid eye movement sleep. Moreover, recent evidence revealed a functional role of aperiodic activity, which reliably discriminates different sleep stages. Aperiodic activity is commonly defined as the spectral slope χ of the 1/frequency (1/fχ) decay function of the electrophysiological power spectrum. However, several lines of inquiry now indicate that the aperiodic component of the power spectrum might be better characterized by a superposition of several decay processes with associated timescales. Here, we determined multiple timescales, which jointly shape aperiodic activity using human intracranial electroencephalography. Across three independent studies (47 participants, 23 female), our results reveal that aperiodic activity reliably dissociated sleep stage-dependent dynamics in a regionally specific manner. A principled approach to parametrize aperiodic activity delineated several, spatially and state-specific timescales. Lastly, we employed pharmacological modulation by means of propofol anesthesia to disentangle state-invariant timescales that may reflect physical properties of the underlying neural population from state-specific timescales that likely constitute functional interactions. Collectively, these results establish the presence of multiple intrinsic timescales that define the electrophysiological power spectrum during distinct brain states.

Interpretable Spatio-Temporal Embedding for Brain Structural-Effective Network with Ordinary Differential Equation

The MRI-derived brain network serves as a pivotal instrument in elucidating both the structural and functional aspects of the brain, encompassing the ramifications of diseases and developmental processes. However, prevailing methodologies, often focusing on synchronous BOLD signals from functional MRI (fMRI), may not capture directional influences among brain regions and rarely tackle temporal functional dynamics. In this study, we first construct the brain-effective network via the dynamic causal model. Subsequently, we introduce an interpretable graph learning framework termed Spatio-Temporal Embedding ODE (STE-ODE). This framework incorporates specifically designed directed node embedding layers, aiming at capturing the dynamic inter-play between structural and effective networks via an ordinary differential equation (ODE) model, which characterizes spatial-temporal brain dynamics. Our framework is validated on several clinical phenotype prediction tasks using two independent publicly available datasets (HCP and OASIS). The experimental results clearly demonstrate the advantages of our model compared to several state-of-the-art methods.

Analyzing asymmetry in brain hierarchies with a linear state-space model of resting-state fMRI data

This study challenges the traditional focus on zero-lag statistics in resting-state functional magnetic resonance imaging (rsfMRI) research. Instead, it advocates for considering time-lag interactions to unveil the directionality and asymmetries of the brain hierarchy. Effective connectivity (EC), the state matrix in dynamical causal modeling (DCM), is a commonly used metric for studying dynamical properties and causal interactions within a linear state-space system description. Here, we focused on how time-lag statistics are incorporated within the framework of DCM resulting in an asymmetric EC matrix. Our approach involves decomposing the EC matrix, revealing a steady-state differential cross-covariance matrix that is responsible for modeling information flow and introducing time-irreversibility. Specifically, the system's dynamics, influenced by the off-diagonal part of the differential covariance, exhibit a curl steady-state flow component that breaks detailed balance and diverges the dynamics from equilibrium. Our empirical findings indicate that the EC matrix's outgoing strengths correlate with the flow described by the differential cross covariance, while incoming strengths are primarily driven by zero-lag covariance, emphasizing conditional independence over directionality.

High-level cognition is supported by information-rich but compressible brain activity patterns

To efficiently yet reliably represent and process information, our brains need to produce information-rich signals that differentiate between moments or cognitive states, while also being robust to noise or corruption. For many, though not all, natural systems, these two properties are often inversely related: More information-rich signals are less robust, and vice versa. Here, we examined how these properties change with ongoing cognitive demands. To this end, we applied dimensionality reduction algorithms and pattern classifiers to functional neuroimaging data collected as participants listened to a story, temporally scrambled versions of the story, or underwent a resting state scanning session. We considered two primary aspects of the neural data recorded in these different experimental conditions. First, we treated the maximum achievable decoding accuracy across participants as an indicator of the "informativeness" of the recorded patterns. Second, we treated the number of features (components) required to achieve a threshold decoding accuracy as a proxy for the "compressibility" of the neural patterns (where fewer components indicate greater compression). Overall, we found that the peak decoding accuracy (achievable without restricting the numbers of features) was highest in the intact (unscrambled) story listening condition. However, the number of features required to achieve comparable classification accuracy was also lowest in the intact story listening condition. Taken together, our work suggests that our brain networks flexibly reconfigure according to ongoing task demands and that the activity patterns associated with higher-order cognition and high engagement are both more informative and more compressible than the activity patterns associated with lower-order tasks and lower engagement.

Increasing the representation of minoritized youth for inclusive and reproducible brain-behavior associations

Population neuroscience datasets allow researchers to estimate reliable effect sizes for brain-behavior associations because of their large sample sizes. However, these datasets undergo strict quality control to mitigate sources of noise, such as head motion. This practice often excludes a disproportionate number of minoritized individuals. We employ motion-ordering and motion-ordering+resampling (bagging) to test if these methods preserve functional MRI (fMRI) data in the Adolescent Brain Cognitive Development Study ( N = 5,733 ). Black and Hispanic youth exhibited excess head motion relative to data collected from White youth, and were discarded disproportionately when using conventional approaches. Both methods retained more than 99% of Black and Hispanic youth. They produced reproducible brain-behavior associations across low-/high-motion racial/ethnic groups based on motion-limited fMRI data. The motion-ordering and bagging methods are two feasible approaches that can enhance sample representation for testing brain-behavior associations and fulfill the promise of consortia datasets to produce generalizable effect sizes across diverse populations.

Spatial transcriptomic patterns underlying amyloid-β and tau pathology are associated with cognitive dysfunction in Alzheimer's disease

Amyloid-β (Aβ) and tau proteins accumulate within distinct neuronal systems in Alzheimer's disease (AD). Although it is not clear why certain brain regions are more vulnerable to Aβ and tau pathologies than others, gene expression may play a role. We study the association between brain-wide gene expression profiles and regional vulnerability to Aβ (gene-to-Aβ associations) and tau (gene-to-tau associations) pathologies by leveraging two large independent AD cohorts. We identify AD susceptibility genes and gene modules in a gene co-expression network with expression profiles specifically related to regional vulnerability to Aβ and tau pathologies in AD. In addition, we identify distinct biochemical pathways associated with the gene-to-Aβ and the gene-to-tau associations. These findings may explain the discordance between regional Aβ and tau pathologies. Finally, we propose an analytic framework, linking the identified gene-to-pathology associations to cognitive dysfunction in AD at the individual level, suggesting potential clinical implication of the gene-to-pathology associations.

Phantom oscillations in principal component analysis

Principal component analysis (PCA) is a dimensionality reduction method that is known for being simple and easy to interpret. Principal components are often interpreted as low-dimensional patterns in high-dimensional space. However, this simple interpretation fails for timeseries, spatial maps, and other continuous data. In these cases, nonoscillatory data may have oscillatory principal components. Here, we show that two common properties of data cause oscillatory principal components: smoothness and shifts in time or space. These two properties implicate almost all neuroscience data. We show how the oscillations produced by PCA, which we call "phantom oscillations," impact data analysis. We also show that traditional cross-validation does not detect phantom oscillations, so we suggest procedures that do. Our findings are supported by a collection of mathematical proofs. Collectively, our work demonstrates that patterns which emerge from high-dimensional data analysis may not faithfully represent the underlying data.

Quantifying the contribution of subject and group factors in brain activation

Research in neuroscience often assumes universal neural mechanisms, but increasing evidence points toward sizeable individual differences in brain activations. What remains unclear is the extent of the idiosyncrasy and whether different types of analyses are associated with different levels of idiosyncrasy. Here we develop a new method for addressing these questions. The method consists of computing the within-subject reliability and subject-to-group similarity of brain activations and submitting these values to a computational model that quantifies the relative strength of group- and subject-level factors. We apply this method to a perceptual decision-making task (n = 50) and find that activations related to task, reaction time, and confidence are influenced equally strongly by group- and subject-level factors. Both group- and subject-level factors are dwarfed by a noise factor, though higher levels of smoothing increases their contributions relative to noise. Overall, our method allows for the quantification of group- and subject-level factors of brain activations and thus provides a more detailed understanding of the idiosyncrasy levels in brain activations.

Living on the edge: network neuroscience beyond nodes

Network neuroscience has emphasized the connectional properties of neural elements - cells, populations, and regions. This has come at the expense of the anatomical and functional connections that link these elements to one another. A new perspective - namely one that emphasizes 'edges' - may prove fruitful in addressing outstanding questions in network neuroscience. We highlight one recently proposed 'edge-centric' method and review its current applications, merits, and limitations. We also seek to establish conceptual and mathematical links between this method and previously proposed approaches in the network science and neuroimaging literature. We conclude by presenting several avenues for future work to extend and refine existing edge-centric analysis.

The biological role of local and global fMRI BOLD signal variability in human brain organization

Variability drives the organization and behavior of complex systems, including the human brain. Understanding the variability of brain signals is thus necessary to broaden our window into brain function and behavior. Few empirical investigations of macroscale brain signal variability have yet been undertaken, given the difficulty in separating biological sources of variance from artefactual noise. Here, we characterize the temporal variability of the most predominant macroscale brain signal, the fMRI BOLD signal, and systematically investigate its statistical, topographical and neurobiological properties. We contrast fMRI acquisition protocols, and integrate across histology, microstructure, transcriptomics, neurotransmitter receptor and metabolic data, fMRI static connectivity, and empirical and simulated magnetoencephalography data. We show that BOLD signal variability represents a spatially heterogeneous, central property of multi-scale multi-modal brain organization, distinct from noise. Our work establishes the biological relevance of BOLD signal variability and provides a lens on brain stochasticity across spatial and temporal scales.

Hierarchical organization of spontaneous co-fluctuations in densely sampled individuals using fMRI

Edge time series decompose functional connectivity into its framewise contributions. Previous studies have focused on characterizing the properties of high-amplitude frames (time points when the global co-fluctuation amplitude takes on its largest value), including their cluster structure. Less is known about middle- and low-amplitude co-fluctuations (peaks in co-fluctuation time series but of lower amplitude). Here, we directly address those questions, using data from two dense-sampling studies: the MyConnectome project and Midnight Scan Club. We develop a hierarchical clustering algorithm to group peak co-fluctuations of all magnitudes into nested and multiscale clusters based on their pairwise concordance. At a coarse scale, we find evidence of three large clusters that, collectively, engage virtually all canonical brain systems. At finer scales, however, each cluster is dissolved, giving way to increasingly refined patterns of co-fluctuations involving specific sets of brain systems. We also find an increase in global co-fluctuation magnitude with hierarchical scale. Finally, we comment on the amount of data needed to estimate co-fluctuation pattern clusters and implications for brain-behavior studies. Collectively, the findings reported here fill several gaps in current knowledge concerning the heterogeneity and richness of co-fluctuation patterns as estimated with edge time series while providing some practical guidance for future studies.

Alterations in static and dynamic regional homogeneity in mesial temporal lobe epilepsy with and without initial precipitating injury

Objectives

Initial precipitating injury (IPI) such as febrile convulsion and intracranial infection will increase the susceptibility to epilepsy. It is still unknown if the functional deficits differ between mesial temporal lobe epilepsy with IPI (mTLE-IPI) and without IPI (mTLE-NO).

Methods

We recruited 25 mTLE-IPI patients, 35 mTLE-NO patients and 33 healthy controls (HC). Static regional homogeneity (sReHo) and dynamic regional homogeneity (dReHo) were then adopted to estimate the alterations of local neuronal activity. One-way analysis of variance was used to analyze the differences between the three groups in sReHo and dReHo. Then the results were utilized as masks for further between-group comparisons. Besides, correlation analyses were carried out to detect the potential relationships between abnormal regional homogeneity indicators and clinical characteristics.

Results

When compared with HC, the bilateral thalamus and the visual cortex in mTLE-IPI patients showed an increase in both sReHo and variability of dReHo. Besides, mTLE-IPI patients exhibited decreased sReHo in the right cerebellum crus1/crus2, inferior parietal lobule and temporal neocortex. mTLE-NO patients showed decreased sReHo and variability of dReHo in the bilateral temporal neocortex compared with HC. Increased sReHo and variability of dReHo were found in the bilateral visual cortex when mTLE-IPI patients was compared with mTLE-NO patients, as well as increased variability of dReHo in the left thalamus and decreased sReHo in the right dorsolateral prefrontal cortex. Additionally, we discovered a negative correlation between the national hospital seizure severity scale testing score and sReHo in the right cerebellum crus1 in mTLE-IPI patients.

Conclusion

According to the aforementioned findings, both mTLE-IPI and mTLE-NO patients had significant anomalies in local neuronal activity, although the functional deficits were much severer in mTLE-IPI patients. The use of sReHo and dReHo may provide a novel insight into the impact of the presence of IPI on the development of mTLE.

Unbiased and efficient sampling of timeseries reveals redundancy of brain network and gradient structure

Many studies in human neuroscience seek to understand the structure of brain networks and gradients. Few studies, however, have tested the redundancy between these outwardly distinct features. Here, we developed methods to directly enable such tests. We built on insights from linear algebra to develop methods for unbiased and efficient sampling of timeseries with network or gradient constraints. We used these methods to show considerable redundancy between popular definitions of network and gradient structure in functional MRI data. On the one hand, we found that network constraints largely accounted for the structure of three major gradients. On the other hand, we found that gradient constraints largely accounted for the structure of seven major networks. Our results imply that some networks and gradients may denote discrete and continuous representations of the same aspects of functional MRI data. We suggest that integrated explanations can reduce redundancy by avoiding the attribution of independent existence or function to these features.

Self-supervised pretraining improves the performance of classification of task functional magnetic resonance imaging

Introduction

Decoding brain activities is one of the most popular topics in neuroscience in recent years. And deep learning has shown high performance in fMRI data classification and regression, but its requirement for large amounts of data conflicts with the high cost of acquiring fMRI data.

Methods

In this study, we propose an end-to-end temporal contrastive self-supervised learning algorithm, which learns internal spatiotemporal patterns within fMRI and allows the model to transfer learning to datasets of small size. For a given fMRI signal, we segmented it into three sections: the beginning, middle, and end. We then utilized contrastive learning by taking the end-middle (i.e., neighboring) pair as the positive pair, and the beginning-end (i.e., distant) pair as the negative pair.

Results

We pretrained the model on 5 out of 7 tasks from the Human Connectome Project (HCP) and applied it in a downstream classification of the remaining two tasks. The pretrained model converged on data from 12 subjects, while a randomly initialized model required 100 subjects. We then transferred the pretrained model to a dataset containing unpreprocessed whole-brain fMRI from 30 participants, achieving an accuracy of 80.2 ± 4.7%, while the randomly initialized model failed to converge. We further validated the model's performance on the Multiple Domain Task Dataset (MDTB), which contains fMRI data of 26 tasks from 24 participants. Thirteen tasks of fMRI were selected as inputs, and the results showed that the pre-trained model succeeded in classifying 11 of the 13 tasks. When using the 7 brain networks as input, variations of the performance were observed, with the visual network performed as well as whole brain inputs, while the limbic network almost failed in all 13 tasks.

Discussion

Our results demonstrated the potential of self-supervised learning for fMRI analysis with small datasets and unpreprocessed data, and for analysis of the correlation between regional fMRI activity and cognitive tasks.