BrainGB: A Benchmark for Brain Network Analysis With Graph Neural Networks

IEEE Transactions on Medical Imaging, Год журнала: 2022, Номер 42(2), С. 493 - 506

Опубликована: Ноя. 4, 2022

Mapping the connectome of human brain using structural or functional connectivity has become one most pervasive paradigms for neuroimaging analysis. Recently, Graph Neural Networks (GNNs) motivated from geometric deep learning have attracted broad interest due to their established power modeling complex networked data. Despite superior performance in many fields, there not yet been a systematic study how design effective GNNs network To bridge this gap, we present BrainGB, benchmark analysis with GNNs. BrainGB standardizes process by (1) summarizing construction pipelines both and modalities (2) modularizing implementation GNN designs. We conduct extensive experiments on datasets across cohorts recommend set general recipes designs networks. support open reproducible research GNN-based analysis, host website at https://braingb.us models, tutorials, examples, as well an out-of-box Python package. hope that work will provide useful empirical evidence offer insights future novel promising direction.

Язык: Английский

---

Neuroscience Needs Network Science

Journal of Neuroscience, Год журнала: 2023, Номер 43(34), С. 5989 - 5995

Опубликована: Авг. 23, 2023

The brain is a complex system comprising myriad of interacting elements, posing significant challenges in understanding its structure, function, and dynamics. Network science has emerged as powerful tool for studying such intricate systems, offering framework integrating multiscale data complexity. Here, we discuss the application network study brain, addressing topics models metrics, connectome, role dynamics neural networks. We explore opportunities multiple streams transitions from development to healthy function disease, potential collaboration between neuroscience communities. underscore importance fostering interdisciplinary through funding initiatives, workshops, conferences, well supporting students postdoctoral fellows with interests both disciplines. By uniting communities, can develop novel network-based methods tailored circuits, paving way towards deeper functions.

Язык: Английский

---

Global Topological Synchronization on Simplicial and Cell Complexes

Physical Review Letters, Год журнала: 2023, Номер 130(18)

Опубликована: Май 3, 2023

Topological signals, i.e., dynamical variables defined on nodes, links, triangles, etc. of higher-order networks, are attracting increasing attention. However, the investigation their collective phenomena is only at its infancy. Here we combine topology and nonlinear dynamics to determine conditions for global synchronization topological signals simplicial or cell complexes. On complexes show that obstruction impedes odd dimensional globally synchronize. other hand, can overcome in some structures any dimension achieve synchronization.

Язык: Английский

---

The dynamic nature of percolation on networks with triadic interactions

Nature Communications, Год журнала: 2023, Номер 14(1)

Опубликована: Март 10, 2023

Percolation establishes the connectivity of complex networks and is one most fundamental critical phenomena for study systems. On simple networks, percolation displays a second-order phase transition; on multiplex transition can become discontinuous. However, little known about in with higher-order interactions. Here, we show that be turned into fully-fledged dynamical process when interactions are taken account. By introducing signed triadic interactions, which node regulate between two other nodes, define percolation. We uncover this paradigmatic model network changes time order parameter undergoes period-doubling route to chaos. provide general theory accurately predicts full diagram random graphs as confirmed by extensive numerical simulations. find real topologies reveals similar phenomenology. These results radically change our understanding may used systems functional changing dynamically non-trivial way, such neural climate networks.

Язык: Английский

---

Dirac signal processing of higher-order topological signals

New Journal of Physics, Год журнала: 2023, Номер 25(9), С. 093013 - 093013

Опубликована: Авг. 23, 2023

Abstract Higher-order networks can sustain topological signals which are variables associated not only to the nodes, but also links, triangles and in general higher dimensional simplices of simplicial complexes. These describe a large variety real systems including currents ocean, synaptic between neurons biological transportation networks. In scenarios signal data might be noisy an important task is process these by improving their noise ratio. So far typically processed independently each other. For instance, node link signals, algorithms that enforce consistent processing across different dimensions largely lacking. Here we propose Dirac processing, adaptive, unsupervised algorithm learns jointly filter supported on links complexes way. The proposed formulated terms discrete operator interpreted as ‘square root’ higher-order Hodge Laplacian. We discuss detail properties its spectrum chirality eigenvectors adopt this formulate defined test our synthetic drifters ocean find learn efficiently reconstruct true outperforming based exclusively

Язык: Английский

---

The coming decade of digital brain research: A vision for neuroscience at the intersection of technology and computing

Imaging Neuroscience, Год журнала: 2024, Номер 2, С. 1 - 35

Опубликована: Апрель 1, 2024

Abstract In recent years, brain research has indisputably entered a new epoch, driven by substantial methodological advances and digitally enabled data integration modelling at multiple scales—from molecules to the whole brain. Major are emerging intersection of neuroscience with technology computing. This science combines high-quality research, across scales, culture multidisciplinary large-scale collaboration, translation into applications. As pioneered in Europe’s Human Brain Project (HBP), systematic approach will be essential for meeting coming decade’s pressing medical technological challenges. The aims this paper to: develop concept decade digital discuss community large, identify points convergence, derive therefrom scientific common goals; provide framework current future development EBRAINS, infrastructure resulting from HBP’s work; inform engage stakeholders, funding organisations institutions regarding research; address transformational potential comprehensive models artificial intelligence, including machine learning deep learning; outline collaborative that integrates reflection, dialogues, societal engagement on ethical opportunities challenges as part research.

Язык: Английский

---

A unified framework for simplicial Kuramoto models

Chaos An Interdisciplinary Journal of Nonlinear Science, Год журнала: 2024, Номер 34(5)

Опубликована: Май 1, 2024

Simplicial Kuramoto models have emerged as a diverse and intriguing class of describing oscillators on simplices rather than nodes. In this paper, we present unified framework to describe different variants these models, categorized into three main groups: “simple” “Hodge-coupled” “order-coupled” (Dirac) models. Our is based topology discrete differential geometry, well gradient systems frustrations, permits systematic analysis their properties. We establish an equivalence between the simple simplicial model standard pairwise networks under condition manifoldness complex. Then, starting from notion synchronization derive bounds coupling strength necessary or sufficient for achieving it. For some variants, generalize results provide new ones, such controllability equilibrium solutions. Finally, explore potential application in reconstruction brain functional connectivity structural connectomes find that edge-based perform competitively even outperform complex extensions node-based

Язык: Английский

---

Information Propagation in Multilayer Systems with Higher-Order Interactions across Timescales

Physical Review X, Год журнала: 2024, Номер 14(2)

Опубликована: Апрель 8, 2024

Complex systems are characterized by multiple spatial and temporal scales. A natural framework to capture their multiscale nature is that of multilayer networks, where different layers represent distinct physical processes often regulate each other indirectly. We model these regulatory mechanisms through triadic higher-order interactions between nodes edges. In this work, we focus on how the timescales associated with layer impact reciprocal effective couplings. First, rigorously derive a decomposition joint probability distribution any dynamical process acting such networks. By inspecting probabilistic structure, unravel general principles governing information propagates across timescales, elucidating interplay mutual causality in systems. particular, show feedback interactions, i.e., those representing from slow fast variables, generate layers. On contrary, direct layers, can propagate only under certain conditions depend solely structure underlying introduce matrix for observables emergent functional apply our results study archetypal examples biological signaling networks environmental dependencies stochastic processes. Our generalizes dynamics paving way deeper understanding real-world shapes content complexity. Published American Physical Society 2024

Язык: Английский

---

The time-evolving epileptic brain network: concepts, definitions, accomplishments, perspectives

Frontiers in Network Physiology, Год журнала: 2024, Номер 3

Опубликована: Янв. 16, 2024

Epilepsy is now considered a network disease that affects the brain across multiple levels of spatial and temporal scales. The paradigm shift from an epileptic focus—a discrete cortical area which seizures originate—to widespread network—spanning lobes hemispheres—considerably advanced our understanding epilepsy continues to influence both research clinical treatment this multi-faceted high-impact neurological disorder. network, however, not static but evolves in time requires novel approaches for in-depth characterization. In review, we discuss conceptual basics theory critically examine state-of-the-art recording techniques analysis tools used assess characterize time-evolving human network. We give account on current shortcomings highlight potential developments towards improved management epilepsy.

Язык: Английский

---

Visual cortical networks for “What” and “Where” to the human hippocampus revealed with dynamical graphs

Cerebral Cortex, Год журнала: 2025, Номер 35(5)

Опубликована: Май 1, 2025

Abstract Key questions for understanding hippocampal function in memory and navigation humans are the type source of visual information that reaches human hippocampus. We measured bidirectional pairwise effective connectivity with functional magnetic resonance imaging between 360 cortical regions while 956 Human Connectome Project participants viewed scenes, faces, tools, or body parts. developed a method using deterministic dynamical graphs to define whole networks flow both directions their over timesteps after signal is applied V1. revealed ventromedial “Where” network from V1 via retrosplenial medial parahippocampal scene areas hippocampus when scenes viewed. A ventrolateral “What” V2–V4, fusiform face cortex, lateral region TF faces/objects There major implications computations vs rodent navigation: primates fovea highly processing process about location objects, landmarks whereas rodents representations system mainly place where individual located self-motion places.

Язык: Английский

---