Multi-Scale Spatio-Temporal Fusion With Adaptive Brain Topology Learning for fMRI Based Neural Decoding DOI
Ziyu Li, Qing Li, Zhiyuan Zhu

и другие.

IEEE Journal of Biomedical and Health Informatics, Год журнала: 2023, Номер 28(1), С. 262 - 272

Опубликована: Окт. 23, 2023

Neural decoding aims to extract information from neurons' activities reveal how the brain functions. Due inherent spatial and temporal characteristics of signals, spatio-temporal computing has become a hot topic for neural decoding. However, extant methods usually use static topology, ignoring dynamic patterns interaction between regions. Further, they do not identify hierarchical organization leading only superficial insight into interactions. Therefore, here we propose novel framework, Multi-Scale Spatio-Temporal framework with Adaptive Brain Topology Learning (MSST-ABTL), It includes two new capabilities enhance decoding: i) ABTL module, which learns topology while updating specific regions, ii) MSST captures association pattern evolution, further enhances interpretability learned multi-scale perspective. We evaluated on public Human Connectome Project (HCP) dataset (resting-state task-related fMRI data). The extensive experiments show that proposed MSST-ABTL outperforms state-of-the-art four evaluation metrics, also can renew neuroscientific discoveries in brain's patterns.

Язык: Английский

Living on the edge: network neuroscience beyond nodes DOI Creative Commons
Richard F. Betzel, Joshua Faskowitz, Olaf Sporns

и другие.

Trends in Cognitive Sciences, Год журнала: 2023, Номер 27(11), С. 1068 - 1084

Опубликована: Сен. 15, 2023

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

Язык: Английский

Процитировано

22
ROSE: A neurocomputational architecture for syntax DOI
Elliot Murphy

Journal of Neurolinguistics, Год журнала: 2023, Номер 70, С. 101180 - 101180

Опубликована: Ноя. 21, 2023

Язык: Английский

Процитировано

16
The biological role of local and global fMRI BOLD signal variability in human brain organization DOI Creative Commons
Giulia Baracchini,

Yigu Zhou,

Jason da Silva Castanheira

и другие.

bioRxiv (Cold Spring Harbor Laboratory), Год журнала: 2023, Номер unknown

Опубликована: Окт. 23, 2023

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

Язык: Английский

Процитировано

15
Unravelling consciousness and brain function through the lens of time, space, and information DOI Creative Commons
Andrea I. Luppi, Fernando Rosas, Pedro A. M. Mediano

и другие.

Trends in Neurosciences, Год журнала: 2024, Номер 47(7), С. 551 - 568

Опубликована: Май 31, 2024

Disentangling how cognitive functions emerge from the interplay of brain dynamics and network architecture is among major challenges that neuroscientists face. Pharmacological pathological perturbations consciousness provide a lens to investigate these complex challenges. Here, we review recent advances about brain's functional organisation have been driven by common denominator: decomposing function into fundamental constituents time, space, information. Whereas unconsciousness increases structure-function coupling across scales, psychedelics may decouple structure. Convergent effects also emerge: anaesthetics, psychedelics, disorders can exhibit similar reconfigurations unimodal-transmodal axis. Decomposition approaches reveal potential translate discoveries species, with computational modelling providing path towards mechanistic integration.

Язык: Английский

Процитировано

6
A method for estimating dynamic functional network connectivity gradients (dFNG) from ICA captures smooth inter-network modulation. DOI Creative Commons
Najme Soleimani, Armin Iraji, Theo G.M. van Erp

и другие.

bioRxiv (Cold Spring Harbor Laboratory), Год журнала: 2024, Номер unknown

Опубликована: Март 11, 2024

Dynamic functional network connectivity (dFNC) analysis is a widely used approach for studying brain function and offering insight into how networks evolve over time. Typically, dFNC studies utilized fixed spatial maps evaluate transient changes in coupling among time courses estimated from independent component (ICA). This manuscript presents complementary that relaxes this assumption by spatially reordering the components dynamically at each timepoint to optimize smooth gradient FNC (i.e., ICA values). Several methods are presented summarize dynamic gradients (dFNGs) time, starting with static (sFNGs), then exploring properties as well dynamics of themselves. We apply dataset schizophrenia (SZ) patients healthy controls (HC). Functional dysconnectivity between different regions has been reported schizophrenia, yet neural mechanisms behind it remain elusive. Using resting state fMRI on consisting 151 160 age gender-matched controls, we extracted 53 intrinsic (ICNs) subject using fully automated constrained approach. develop several summaries our analysis, both sense, computed Pearson correlation coefficient full series, sliding window followed based gradient, group differences. Static revealed significantly stronger subcortical (SC), auditory (AUD) visual (VIS) patients, hypoconnectivity sensorimotor (SM) relative controls. sFNG highlighted distinctive clustering patterns HCs along cognitive control (CC)/ default mode (DMN), SC/ AUD/ SM/ cerebellar (CB), VIS gradients. Furthermore, observed significant differences sFNGs groups SC CB domains. dFNG suggested SZ spend more first while favor SM/DMN state. For second however, exhibited higher activity domains, contrasting HCs' DMN engagement. The synchrony conveyed shifts transmodal CC/ patients. In addition, distinct SC, SM domains compared HCs. To recap, results advance understanding modulation examining trajectories. provides complete spatiotemporal summary data, contributing growing body current literature regarding By employing dFNG, highlight new perspective capture large scale fluctuations across maintaining convenience low dimensional measures.

Язык: Английский

Процитировано

5
Complexity and entropy of natural patterns DOI Creative Commons

H Wang,

Changqing Song, Peichao Gao

и другие.

PNAS Nexus, Год журнала: 2024, Номер 3(10)

Опубликована: Сен. 19, 2024

Complexity and entropy play crucial roles in understanding dynamic systems across various disciplines. Many intuitively perceive them as distinct measures assume that they have a concave-down relationship. In everyday life, there is common consensus while never decreases, complexity does decrease after an initial increase during the process of blending coffee milk. However, this primarily conceptual lacks empirical evidence. Here, we provide comprehensive evidence challenges prevailing consensus. We demonstrate is, fact, illusion resulting from choice system characterization (dimension) unit observation (resolution). By employing measure designed for natural patterns, find coffee-milk decreases if appropriately characterized terms dimension resolution. Also, aligns experimentally theoretically with entropy, suggesting it not represent so-called effective complexity. These findings rectify reshape our relationship between entropy. It therefore to exercise caution pay close attention accurately precisely characterize before delving into their underlying mechanisms, despite maturity research fields dealing patterns such geography ecology. The characterization/observation (dimension resolution) fundamentally determines assessment using existing understanding.

Язык: Английский

Процитировано

4
State-dependent signatures of anti-N-methyl-d-aspartate receptor encephalitis DOI Creative Commons
Nina von Schwanenflug, Stephan Krohn, J. Heine

и другие.

Brain Communications, Год журнала: 2022, Номер 4(1)

Опубликована: Янв. 3, 2022

Traditional static functional connectivity analyses have shown distinct network alterations in patients with anti-

Язык: Английский

Процитировано

16
Functional connectivity dynamics reflect disability and multi-domain clinical impairment in patients with relapsing-remitting multiple sclerosis DOI Creative Commons
Amy Romanello, Stephan Krohn, Nina von Schwanenflug

и другие.

NeuroImage Clinical, Год журнала: 2022, Номер 36, С. 103203 - 103203

Опубликована: Янв. 1, 2022

Multiple sclerosis (MS) is an autoimmune disease of the central nervous system associated with deficits in cognitive and motor functioning. While structural brain changes such as demyelination are early hallmark disease, a characteristic profile functional alterations MS lacking. Functional neuroimaging studies at various stages have revealed complex heterogeneous patterns aberrant connectivity (FC) MS, previous largely being limited to static account FC. Thus, it remains unclear how time-resolved FC relates variance clinical disability status MS. We here aimed characterize network organization patients analysis explore relationship between status, multi-domain outcomes altered dynamics. Resting-state MRI (rs-fMRI) data were acquired from 101 age- sex-matched healthy controls (HC). Based on Expanded Disability Status Score (EDSS), split into two sub-groups: without (EDSS≤1, n = 36) mild moderate levels (EDSS≥2, 39). Five dynamic states extracted whole-brain rs-fMRI data. Group differences strength, across-state overall connectivity, dwell time, transition frequency, modularity, global assessed. Patients' impairment was quantified custom outcome z-scores (higher: worse) for domains depressive symptoms, fatigue, motor, vision, cognition, total atrophy, lesion load. Correlation analyses measures performed Spearman partial correlation controlling age. Patients exhibited more widespread spatiotemporal pattern spent time high-connectivity, low-occurrence state compared HCs. Worse symptoms all positively EDSS scores. Furthermore, symptom severity related dynamics measured by state-specific DMN attention networks, while fatigue reduced frontoparietal basal ganglia. Despite comparably low we identified distinct those disability, these sensitive multiple domains. uncovered correlations that remained undetected conventional analyses, showing accounting temporal helps disentangle alterations,

Язык: Английский

Процитировано

16
Reliability of variability and complexity measures for task and task‐free BOLD fMRI DOI Creative Commons
Maren H. Wehrheim, Joshua Faskowitz, Anna‐Lena Schubert

и другие.

Human Brain Mapping, Год журнала: 2024, Номер 45(10)

Опубликована: Июль 9, 2024

Abstract Brain activity continuously fluctuates over time, even if the brain is in controlled (e.g., experimentally induced) states. Recent years have seen an increasing interest understanding complexity of these temporal variations, for example with respect to developmental changes function or between‐person differences healthy and clinical populations. However, psychometric reliability signal variability measures—which important precondition robust individual as well longitudinal research—is not yet sufficiently studied. We examined (split‐half correlations) test–retest correlations task‐free (resting‐state) BOLD fMRI split‐half seven functional task data sets from Human Connectome Project evaluate their reliability. observed good excellent measures derived rest activation time series (standard deviation, mean absolute successive difference, squared difference), moderate same under conditions. estimates (several entropy dimensionality measures) showed reliabilities both, calculated also time‐resolved (dynamic) connectivity measures, but poor series. Global (i.e., across cortical regions) tended show higher than region‐specific estimates. Larger subcortical regions similar regions, small lower reliability, especially measures. Lastly, we that scores are only minorly dependent on scan length replicate our results different parcellation denoising strategies. These suggest well‐suited research. Temporal global provides novel approach robustly quantifying dynamics function. Practitioner Points Variability Measures can quantify neural dynamics. Length has a minor effect

Язык: Английский

Процитировано

3
Identifying Distinct Developmental Patterns of Brain Complexity in Autism: A Cross‐Sectional Cohort Analysis Using the Autism Brain Imaging Data Exchange DOI Creative Commons
I‐Jou Chi, Shih‐Jen Tsai, Chun‐Houh Chen

и другие.

Psychiatry and Clinical Neurosciences, Год журнала: 2025, Номер unknown

Опубликована: Янв. 11, 2025

Aim Autistic traits exhibit neurodiversity with varying behaviors across developmental stages. Brain complexity theory, illustrating the dynamics of neural activity, may elucidate evolution autistic over time. Our study explored patterns brain in individuals from childhood to adulthood. Methods We analyzed functional magnetic resonance imaging data 1087 participants and neurotypical controls aged 6 30 years within ABIDE I (Autism Imaging Data Exchange) set. Sample entropy was calculated measure among 90 regions, utilizing an automated anatomical labeling template for voxel parcellation. Participants were grouped using sliding age windows partial overlaps. assessed average entire regions both groups categories. Cluster analysis conducted generalized association plots identify similar trajectories. Finally, relationship between region examined. Results tend toward higher whole‐brain during adolescence lower adulthood, indicating possible distinct However, these results do not remain after Bonferroni correction. Two clusters identified, each unique changes Correlations complexity, age, also identified. Conclusion The revealed trajectories individuals, providing insight into autism suggesting that age‐related could be a potential neurodevelopmental marker dynamic nature autism.

Язык: Английский

Процитировано

0