Pith. sign in

REVIEW 4 major objections 6 minor 151 references

Hyperbolic geometry maps brain hierarchy, boosting disorder diagnosis

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

T0 review · glm-5.2

2026-07-09 20:29 UTC pith:6B45P47G

load-bearing objection GaMamba is the real contribution; the hyperbolic module doesn't clearly earn its keep the 4 major comments →

arxiv 2607.07077 v1 pith:6B45P47G submitted 2026-07-08 cs.CV cs.AI

Navigating Hierarchy: Hyperbolic Learning on Brain Graphs for Disorder Diagnosis

Yapeng Li , Bo Jiang , Ziyan Zhang , Dongdong Chen , Zhengzheng Tu This is my paper
classification cs.CV cs.AI
keywords brainhyperbolicnetworkhierarchicalhlbglearningwhole-brainfunctional
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper argues that brain networks have a natural three-tier hierarchy—individual regions, functional communities, and the whole brain—and that this hierarchy is best captured not in flat Euclidean space but in hyperbolic space, a geometry with constant negative curvature that naturally accommodates tree-like, nested structures. The authors propose a framework called HLBG that projects region, community, and whole-brain representations into Lorentzian hyperbolic space and then enforces two geometric entailment constraints: one requiring each region to fall within the entailment cone of its parent community, and another requiring each community to fall within the cone of the whole brain. This forces the learned representations to respect the part-to-whole containment structure of functional brain organization. The paper also introduces a Graph-aware Mamba module that injects graph topological structure into the Mamba state-space model, allowing it to capture long-range dependencies across brain regions while preserving local connectivity patterns. The combined system is tested on autism (ABIDE-I) and depression (REST-MDD) classification, yielding modest but consistent accuracy improvements over Euclidean graph neural networks and graph Transformers, and producing attention-based biomarker maps that align with known neuroscience findings.

Core claim

The central claim is that explicitly imposing ROI-to-community-to-whole-brain hierarchical containment constraints in Lorentzian hyperbolic space produces more discriminative brain-network representations than flat-space methods, because the negative-curvature geometry naturally mirrors the nested, tree-like organization of functional brain architecture. The entailment-cone loss is the load-bearing mechanism: it penalizes region representations that fall outside their community's cone and community representations that fall outside the whole-brain cone, thereby enforcing a geometric hierarchy that Euclidean aggregation cannot express.

What carries the argument

The two entailment losses are the novel geometric machinery. Each parent node (community or whole-brain) defines an entailment cone in hyperbolic space—a region of the manifold whose aperture depends on the parent's distance from the origin. The loss function penalizes any child whose angular position relative to the parent exceeds the cone's half-aperture angle. This converts the abstract notion of hierarchy into a concrete geometric constraint: children must lie within their parent's cone, which in hyperbolic space is exponentially cheaper to satisfy at deeper tree levels than in Euclidean space. The second mechanism is GaMamba, which modifies Mamba's output readout matrix by adding a GAT-

Load-bearing premise

The entailment-cone constraint assumes that the functional brain hierarchy is strict enough that each region belongs unambiguously within a single community's cone and each community within the whole-brain cone. In reality, many brain regions—hub regions like the precuneus—participate in multiple overlapping functional networks simultaneously, so the true structure may be a graph with cross-links rather than a clean tree.

What would settle it

If one were to take brain regions known to participate in multiple communities (e.g., the precuneus or medial prefrontal cortex) and measure their angular distance to each candidate parent community in the learned hyperbolic embedding, a strict entailment model would predict that they fall within exactly one cone. If instead these regions are systematically pushed to the boundaries or outside of all community cones, or if the model's classification accuracy degrades specifically for subjects whose functional architecture is most non-tree-like, that would indicate the strict hierarchy is distal

Watch this falsifier. Get emailed when new claim-graph text bears on it.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit.

Referee Report

4 major / 6 minor

Summary. The manuscript proposes HLBG, a framework for brain disorder diagnosis from fMRI functional connectivity graphs. The method has two main contributions: (1) a Graph-aware Mamba (GaMamba) module that injects GAT-derived structural prompts into the selective state-space model's readout matrix, and (2) a Hierarchical Brain Representation Learning (HBRL) module that projects ROI-, community-, and whole-brain-level embeddings into Lorentzian hyperbolic space and enforces ROI→community→whole-brain entailment via two geometric cone losses. Experiments on ABIDE-I (ASD) and REST-MDD (MDD) show improvements over GNN, Graph Transformer, and Mamba baselines. The mathematical formulations follow standard hyperbolic geometry and SSM definitions. The core concern is that the central contribution—hyperbolic hierarchical modeling—is not clearly supported by the ablation evidence, and headline gains lack statistical testing.

Significance. The paper addresses a genuine gap: most brain-graph methods model either flat node-level or community-level structure without an explicit geometric hierarchy. The idea of using Lorentzian entailment cones to enforce ROI⊂community⊂brain relationships is novel for this application. The GaMamba design—injecting topology into the SSM readout rather than relying solely on node ordering—is a reasonable architectural choice. However, the significance is tempered by the fact that the hyperbolic component shows inconsistent ablation benefits, and the empirical gains over the strongest baseline fall within the reported standard-deviation envelope without statistical testing.

major comments (4)
  1. Table 3 (ablation): The HBRL module—the paper's central contribution—shows inconsistent or marginal ablation effects. On REST-MDD, adding HBRL to GaMamba alone *decreases* ACC from 69.24% to 68.66% and AUC from 73.35% to 72.60%. On ABIDE-I, HBRL adds only 0.49% ACC over GaMamba alone (73.91→74.40). The full model's gains over the second-best baseline (CAGT) are 1.44% ACC on ABIDE-I and 1.90% on REST-MDD, but the reported standard deviations are ±3.32 and ±2.27 respectively. No paired statistical test (paired t-test, Wilcoxon signed-rank, etc.) is reported across the 10 folds. The authors should either (a) provide paired statistical tests across folds to demonstrate that the improvements are significant, or (b) acknowledge that the hyperbolic component's contribution is marginal and reframe the central claim accordingly. As it stands, the empirical evidence does not clearly establish that
  2. Table 2 (comparison with SOTA): The headline improvements over the second-best method (CAGT) are 1.44% ACC on ABIDE-I and 1.90% on REST-MDD. Given the standard deviations (±3.32 and ±2.27), these differences are within the noise envelope of 10-fold cross-validation. Without a paired statistical test comparing HLBG against CAGT fold-by-fold, it is not possible to determine whether the reported gains are real or artifacts of variance. The authors should report per-fold results for at least the top two methods and apply an appropriate paired test.
  3. §3.3, Eqs. (9)–(11): The entailment-cone loss enforces a strict tree-like hierarchy where each ROI falls within the entailment cone of exactly one community, and each community within the whole-brain cone. However, brain regions such as the precuneus participate in multiple functional networks simultaneously. The manuscript does not discuss how this strict tree assumption interacts with the overlapping, non-hierarchical structure of functional brain organization. The authors should either (a) discuss this limitation explicitly and justify why the tree approximation is reasonable, or (b) consider a softer entailment that allows partial membership. The concern is not that the tree assumption is wrong by definition, but that without any discussion or sensitivity analysis, it is unclear whether the constraint could distort representations for hub regions.
  4. §3.3, Eq. (5)–(6): The aggregation function φ(·) that produces community-level (Z_c) and whole-brain (Z_b) representations from ROI features is an attention-based weighted summation. This aggregation is performed in Euclidean space *before* the exponential map into hyperbolic space. The entailment losses then operate on the hyperbolic projections of these pre-aggregated vectors. The authors should clarify whether Euclidean aggregation followed by hyperbolic projection is geometrically consistent—i.e., whether the resulting hyperbolic points meaningfully represent the 'parent' of their children in the Lorentz model, or whether the aggregation should be performed using Fréchet means or other hyperbolic-native operations to ensure that the entailment losses.
minor comments (6)
  1. Figure 3 contains garbled characters (e.g., '/uni00000024/uni00000026...') instead of readable axis labels. This should be replaced with legible labels.
  2. §3.1: The top-k parameter for adjacency construction is mentioned (k=30 in §4.2) but the sensitivity to this choice is not analyzed. A brief note on robustness would improve readability.
  3. §4.2: The curvature parameter κ is set to 1.3 for ABIDE-I and 0.39 for REST-MDD, but no justification is given for these different values or how they were selected. A brief explanation would help readers understand whether this is dataset-specific tuning or principled selection.
  4. Several references have future dates (e.g., Wang et al. 2026, Jia et al. 2026). If these are accepted/in-press works, this should be noted; if they are arXiv preprints, the citation format should reflect this.
  5. §3.4: The OCRead module is referenced as following prior work (Kan et al. 2022; Pei et al. 2025) but its role in the overall pipeline is described only briefly. A one-sentence explanation of why structured graph-level readout is preferred over simple pooling would help readers.
  6. Table 2: The 'vanillaTF' baseline (Vaswani et al. 2017) is a generic Transformer applied to brain graphs. It would be useful to clarify how it is adapted for graph input (e.g., using node sequences).

Circularity Check

0 steps flagged

No significant circularity; the hierarchy is predefined but the entailment constraints add genuine geometric structure not implied by the aggregation alone.

full rationale

The paper's derivation chain is self-contained. The hierarchical brain graph (Section 3.1) is constructed using the external Yeo atlas (Yeo et al., 2011) to assign ROIs to communities — this is a modeling assumption, not a circular definition. The HBRL module (Section 3.3) computes community-level features Z_c as attention-based aggregations of constituent ROI features (Eq. 5-6), then projects all levels into Lorentzian hyperbolic space (Eq. 7) and imposes entailment cone constraints (Eq. 9-11). The entailment losses enforce a geometric property (child node lies within parent's entailment cone) that is NOT automatically satisfied by the Euclidean aggregation operation — the aggregation defines WHAT the community representation is, while the entailment loss constrains WHERE it lands in hyperbolic space. These are distinct operations. The exponential map and Lorentzian inner product (Eq. 7-8) follow standard external references (Cannon et al., 1997; Desai et al., 2023; Ganea et al., 2018; Pal et al., 2025), none of which are self-citations by the current authors. The classification pipeline (Section 3.4) uses standard supervised cross-entropy loss plus the entailment regularizers, with no fitted-parameter-renamed-as-prediction pattern. The biomarker analysis (Section 4.6) is post-hoc attention visualization, not a prediction that reduces to inputs. The reader's concern that the hierarchy is 'imposed a priori' rather than 'discovered' is a valid modeling critique but does not constitute circularity: the entailment constraints add information (geometric structure in hyperbolic space) beyond what the aggregation alone provides. Score 1 reflects the mild observation that community representations are defined as aggregations of their constituent ROIs and then constrained to 'entail' those same ROIs, but this is not true circularity since the geometric constraint is not trivially satisfied by the aggregation.

Axiom & Free-Parameter Ledger

7 free parameters · 4 axioms · 2 invented entities

The paper introduces 7 free parameters tuned per dataset, 4 axioms (3 domain assumptions about brain hierarchy and 1 standard mathematical assumption about SSMs), and 2 invented entities (GaMamba and HBRL modules). The domain assumptions about strict brain hierarchy and fixed community parcellation are the most consequential, as they determine whether the hyperbolic entailment constraints are appropriate for the data.

free parameters (7)
  • kappa (curvature) = 1.3 (ABIDE-I), 0.39 (REST-MDD)
    Absolute curvature of the Lorentzian hyperbolic space, set differently per dataset without principled justification.
  • lambda_1 = 0.4 (ABIDE-I), 0.3 (REST-MDD)
    Weight for ROI-in-Community entailment loss, tuned per dataset.
  • lambda_2 = 0.1 (ABIDE-I), 0.3 (REST-MDD)
    Weight for Community-in-Brain entailment loss, tuned per dataset.
  • w = 0.3 (ABIDE-I), 0.1 (REST-MDD)
    Weight coefficient for self-attention fusion of global and local representations.
  • N_b (prompt token size) = 10 (ABIDE-I), 8 (REST-MDD)
    Number of learnable prompt tokens for subgraph-aware attention, tuned per dataset.
  • K (boundary constant) = 0.1
    Constant in the half-aperture angle formula, set to impose boundary condition near the origin.
  • k (top-k connections) = 30
    Number of strongest functional connections retained per node in the adjacency matrix.
axioms (4)
  • domain assumption Functional brain networks exhibit strict hierarchical organization (ROI -> community -> whole-brain) amenable to tree-like entailment constraints.
    The entire HBRL module (Section 3.3) depends on this assumption. Brain regions often participate in multiple overlapping networks, which may violate strict hierarchy.
  • domain assumption The Yeo 7-network parcellation provides a valid community structure for all subjects.
    Section 3.1 assigns ROIs to communities using the standardized Yeo atlas. Individual subject variability in community structure is not modeled.
  • domain assumption Lorentzian hyperbolic space with constant negative curvature is an appropriate geometry for brain-network hierarchy.
    Section 3.3 projects representations into Lorentzian space. The appropriateness of constant-curvature vs. variable-curvature geometry for brain networks is not justified.
  • standard math The selective state-space model (Mamba) can be extended to graph-structured data via node ordering and structural prompt injection.
    Section 3.2 builds GaMamba on the SSM framework. The mathematical formulation follows standard SSM equations with modifications.
invented entities (2)
  • GaMamba module no independent evidence
    purpose: Graph-aware Mamba that injects GAT-derived structural prompts into Mamba's readout matrix for topology-aware long-range dependency modeling.
    The module is evaluated only within the paper's own framework on the paper's own datasets. No external reproduction or independent evaluation is provided.
  • Hierarchical Brain Representation Learning (HBRL) module no independent evidence
    purpose: Projects multi-level brain representations into Lorentzian hyperbolic space and imposes entailment constraints.
    The module's contribution is evaluated only through the paper's ablation studies. The entailment constraints are not validated against alternative hierarchical regularization approaches.

pith-pipeline@v1.1.0-glm · 24792 in / 3089 out tokens · 212561 ms · 2026-07-09T20:29:33.434448+00:00 · methodology

0 comments
read the original abstract

Functional brain networks exhibit a hierarchical organization across ROI, community, and whole-brain levels, supporting local processing, inter-community coordination, and global integration. Recent studies have demonstrated that brain community-aware modeling is beneficial for both diagnosis and biomarker identification of brain networks. However, existing brain graph modeling methods often struggle to model ROI-community interactions, thereby failing to fully exploit the hierarchy across ROI, community, and whole-brain network levels. To address this issue, inspired by deep hyperbolic learning in modeling hierarchical structures, we propose a novel framework, termed Hyperbolic Learning on Brain Graphs (HLBG), for brain network analysis. The core idea of HLBG is to exploit the inherent hierarchical geometry of hyperbolic space to model the hierarchical relationships among ROIs, functional communities, and the whole-brain network, thereby learning hierarchy-aware and highly discriminative representations for brain network data. Specifically, HLBG first projects representations from ROIs, communities, and the whole-brain network into Lorentzian hyperbolic space. Then, the multi-level hierarchy is imposed via two geometric entailment constraints. In addition, we introduce a new Graph-aware Mamba (GaMamba) model, which incorporates topology-derived structural prompts into Mamba to capture long-range dependencies while preserving graph topological information. Experiments on ABIDE-I and REST-MDD demonstrate that HLBG outperforms state-of-the-art methods and identifies disorder-relevant functional biomarkers.

Figures

Figures reproduced from arXiv: 2607.07077 by Bo Jiang, Dongdong Chen, Yapeng Li, Zhengzheng Tu, Ziyan Zhang.

Figure 1
Figure 1. Figure 1: (a) Illustration of the hierarchical organization of the brain graph across the ROI, community, and whole-brain levels. (b) Representation of the hierarchical brain network structure in hyperbolic space. Behrouz and Hashemi, 2024). However, existing Graph￾Mamba models struggle to incorporate graph topology while retaining Mamba’s inherent long-range modeling capability. Recently, some studies have demonstr… view at source ↗
Figure 2
Figure 2. Figure 2: Overall framework of the HLBG method. It mainly includes three modules: Hierarchical Brain Graph Construction; Brain Graph Embedding via Graph-aware Mamba; Hierarchical Brain Representation Learning. GaMamba to characterize topology-aware global depen￾dencies while simultaneously encoding fine-grained intra￾community features. Third, we propose a Hierarchical Brain Representation Learning module to explici… view at source ↗
Figure 3
Figure 3. Figure 3: Hyperparameter sensitivity analysis of 𝜆1 , 𝜆2 , 𝑤 and 𝑁𝑏 . Panels (a)–(d) present results on the ABIDE dataset, while panels (e)–(h) show results on the REST-MDD dataset. Baseline GAT Mamba All 65.0 67.5 70.0 72.5 75.0 77.5 80.0 Performance (%) 72.17 72.65 73.13 75.42 75.45 75.92 78.46 78.97 ABIDE-I ACC AUC Baseline GAT Mamba All 68.19 67.61 68.78 70.13 72.28 70.28 72.45 74.11 REST-MDD ACC AUC [PITH_FULL… view at source ↗
Figure 4
Figure 4. Figure 4: Ablation analysis of the Graph-aware Mamba module. We evaluate these hyperparameters on ABIDE-I and REST￾MDD, with results presented in [PITH_FULL_IMAGE:figures/full_fig_p009_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Visualization of the brain connectivity networks and key brain regions on ABIDE-I and REST-MDD datasets. (a) and (c) visualize the top 30 most critical brain functional connections derived from the attention matrices 𝐀ASD and 𝐀MDD, respectively. (b) and (d) visualize the top ten key ROIs identified from the attention matrices 𝐀ASD and 𝐀MDD, respectively. precuneus- and orbitofrontal-related functional abno… view at source ↗

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

151 extracted references · 151 canonical work pages · 1 internal anchor

  1. [1]

    International Conference on Learning Representations , year=

    Strategies for Pre-training Graph Neural Networks , author=. International Conference on Learning Representations , year=

    work page

  2. [2]

    Medical Image Analysis , volume=

    Braingnn: Interpretable brain graph neural network for fmri analysis , author=. Medical Image Analysis , volume=. 2021 , publisher=

    work page 2021

  3. [3]

    International conference on medical image computing and computer-assisted intervention , pages=

    Interpretable graph neural networks for connectome-based brain disorder analysis , author=. International conference on medical image computing and computer-assisted intervention , pages=. 2022 , organization=

    work page 2022

  4. [4]

    International Conference on Medical Image Computing and Computer-Assisted Intervention , pages=

    Brainusl: Unsupervised graph structure learning for functional brain network analysis , author=. International Conference on Medical Image Computing and Computer-Assisted Intervention , pages=. 2023 , organization=

    work page 2023

  5. [5]

    IEEE Transactions on Neural Networks and Learning Systems , year=

    Brainib: Interpretable brain network-based psychiatric diagnosis with graph information bottleneck , author=. IEEE Transactions on Neural Networks and Learning Systems , year=

    work page

  6. [6]

    Advances in Neural Information Processing Systems , volume=

    Brain network transformer , author=. Advances in Neural Information Processing Systems , volume=

    work page

  7. [7]

    International conference on medical image computing and computer-assisted intervention , pages=

    Community-aware transformer for autism prediction in fmri connectome , author=. International conference on medical image computing and computer-assisted intervention , pages=. 2023 , organization=

    work page 2023

  8. [8]

    Advances in Neural Information Processing Systems , volume=

    Long-range brain graph transformer , author=. Advances in Neural Information Processing Systems , volume=

    work page

  9. [9]

    International Conference on Medical Image Computing and Computer-Assisted Intervention , pages=

    Gbt: Geometric-oriented brain transformer for autism diagnosis , author=. International Conference on Medical Image Computing and Computer-Assisted Intervention , pages=. 2024 , organization=

    work page 2024

  10. [10]

    Proceedings of the 33rd ACM International Conference on Information and Knowledge Management , pages=

    Contrasformer: a brain network contrastive transformer for neurodegenerative condition identification , author=. Proceedings of the 33rd ACM International Conference on Information and Knowledge Management , pages=

    work page

  11. [11]

    Advances in Neural Information Processing Systems , volume=

    Recipe for a general, powerful, scalable graph transformer , author=. Advances in Neural Information Processing Systems , volume=

    work page

  12. [13]

    Human brain mapping , volume=

    A whole brain fMRI atlas generated via spatially constrained spectral clustering , author=. Human brain mapping , volume=. 2012 , publisher=

    work page 2012

  13. [14]

    Proceedings of the Thirty-Fourth International Joint Conference on Artificial Intelligence,

    Community-Aware Graph Transformer for Brain Disorder Identification , author =. Proceedings of the Thirty-Fourth International Joint Conference on Artificial Intelligence,. 2025 , month =

    work page 2025

  14. [15]

    Psychoradiology , volume=

    The DIRECT consortium and the REST-meta-MDD project: towards neuroimaging biomarkers of major depressive disorder , author=. Psychoradiology , volume=. 2022 , publisher=

    work page 2022

  15. [16]

    Proceedings of the National Academy of Sciences , volume=

    Reduced default mode network functional connectivity in patients with recurrent major depressive disorder , author=. Proceedings of the National Academy of Sciences , volume=. 2019 , publisher=

    work page 2019

  16. [17]

    Frontiers in Neuroinformatics , volume=

    The neuro bureau preprocessing initiative: open sharing of preprocessed neuroimaging data and derivatives , author=. Frontiers in Neuroinformatics , volume=

    work page

  17. [18]

    Neuroinformatics , volume=

    DPABI: data processing & analysis for (resting-state) brain imaging , author=. Neuroinformatics , volume=. 2016 , publisher=

    work page 2016

  18. [19]

    Scientific data , volume=

    Standardizing human brain parcellations , author=. Scientific data , volume=. 2021 , publisher=

    work page 2021

  19. [20]

    Noise reduction in speech processing , pages=

    Pearson correlation coefficient , author=. Noise reduction in speech processing , pages=. 2009 , publisher=

    work page 2009

  20. [21]

    IEEE Transactions on Emerging Topics in Computational Intelligence , volume=

    Multi-modal non-euclidean brain network analysis with community detection and convolutional autoencoder , author=. IEEE Transactions on Emerging Topics in Computational Intelligence , volume=. 2022 , publisher=

    work page 2022

  21. [22]

    IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) , pages=

    Brain-hgcn: A hyperbolic graph convolutional network for brain functional network analysis , author=. IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) , pages=. 2026 , organization=

    work page 2026

  22. [23]

    First conference on language modeling , year=

    Mamba: Linear-time sequence modeling with selective state spaces , author=. First conference on language modeling , year=

    work page

  23. [24]

    International Conference on Machine Learning , pages=

    Hyperbolic image-text representations , author=. International Conference on Machine Learning , pages=. 2023 , organization=

    work page 2023

  24. [25]

    Flavors of geometry , volume=

    Hyperbolic geometry , author=. Flavors of geometry , volume=

    work page

  25. [26]

    MNM: Multi-level Neuroimaging Meta-analysis with Hyperbolic Brain-Text Representations

    Baek, Seunghun and Lee, Jaejin and Sim, Jaeyoon and Jeong, Minjae and Kim, Won Hwa. MNM: Multi-level Neuroimaging Meta-analysis with Hyperbolic Brain-Text Representations. Medical Image Computing and Computer Assisted Intervention -- MICCAI 2025. 2026

    work page 2025

  26. [27]

    Compositional Entailment Learning for Hyperbolic Vision-Language Models , volume =

    Avik Pal and Max van Spengler and Guido Maria D'Amely di Melendugno and Alessandro Flaborea and Fabio Galasso and Pascal Mettes , booktitle =. Compositional Entailment Learning for Hyperbolic Vision-Language Models , volume =

    work page

  27. [28]

    International conference on machine learning , pages=

    Hyperbolic entailment cones for learning hierarchical embeddings , author=. International conference on machine learning , pages=. 2018 , organization=

    work page 2018

  28. [29]

    Science , volume=

    Brain disorders? precisely , author=. Science , volume=. 2015 , publisher=

    work page 2015

  29. [30]

    Journal of autism and developmental disorders , volume=

    Severe impairments of social interaction and associated abnormalities in children: Epidemiology and classification , author=. Journal of autism and developmental disorders , volume=. 1979 , publisher=

    work page 1979

  30. [31]

    IEEE transactions on medical imaging , volume=

    Virtual adversarial training-based deep feature aggregation network from dynamic effective connectivity for MCI identification , author=. IEEE transactions on medical imaging , volume=. 2021 , publisher=

    work page 2021

  31. [32]

    IEEE Transactions on Medical Imaging , volume=

    A multi-graph cross-attention-based region-aware feature fusion network using multi-template for brain disorder diagnosis , author=. IEEE Transactions on Medical Imaging , volume=. 2023 , publisher=

    work page 2023

  32. [33]

    International conference on medical image computing and computer-assisted intervention , pages=

    Pooling regularized graph neural network for fmri biomarker analysis , author=. International conference on medical image computing and computer-assisted intervention , pages=. 2020 , organization=

    work page 2020

  33. [34]

    Proceedings of the Thirty-Third International Joint Conference on Artificial Intelligence,

    Graph Neural Networks for Brain Graph Learning: A Survey , author =. Proceedings of the Thirty-Third International Joint Conference on Artificial Intelligence,. 2024 , month =

    work page 2024

  34. [35]

    Biologically Plausible Brain Graph Transformer , volume =

    Peng, Ciyuan and Huang, Yuelong and Dong, Qichao and Yu, Shuo and Xia, Feng and Zhang, Chengqi and Jin, Yaochu , booktitle =. Biologically Plausible Brain Graph Transformer , volume =

    work page

  35. [36]

    Cerebral Cortex , volume=

    DS-GCNs: Connectome classification using dynamic spectral graph convolution networks with assistant task training , author=. Cerebral Cortex , volume=. 2021 , publisher=

    work page 2021

  36. [37]

    IEEE transactions on medical imaging , volume=

    Braingb: a benchmark for brain network analysis with graph neural networks , author=. IEEE transactions on medical imaging , volume=. 2022 , publisher=

    work page 2022

  37. [38]

    Medical Image Analysis , volume=

    Knowledge-driven multi-graph convolutional network for brain network analysis and potential biomarker discovery , author=. Medical Image Analysis , volume=. 2025 , publisher=

    work page 2025

  38. [39]

    Pattern Recognition , volume=

    Multi-scale enhanced graph convolutional network for mild cognitive impairment detection , author=. Pattern Recognition , volume=. 2023 , publisher=

    work page 2023

  39. [40]

    Trifocal Transformer: Connection-Mask-Residual Focused Attention Network for Brain Disease Diagnosis , year=

    Wang, Bin and Liang, Jiarui and Ye, Chuyang and Yan, Ting and Liu, Miaomiao and Yan, Tianyi , journal=. Trifocal Transformer: Connection-Mask-Residual Focused Attention Network for Brain Disease Diagnosis , year=

    work page

  40. [41]

    Proceedings of the 30th ACM SIGKDD Conference on Knowledge Discovery and Data Mining , pages =

    Behrouz, Ali and Hashemi, Farnoosh , title =. Proceedings of the 30th ACM SIGKDD Conference on Knowledge Discovery and Data Mining , pages =. 2024 , isbn =

    work page 2024

  41. [42]

    Multi-view brain network classification based on Adaptive Graph Isomorphic Information Bottleneck Mamba , journal =

    Changxu Dong and Dengdi Sun and Zhenda Yu and Bin Luo , keywords =. Multi-view brain network classification based on Adaptive Graph Isomorphic Information Bottleneck Mamba , journal =. 2025 , issn =. doi:https://doi.org/10.1016/j.eswa.2024.126170 , url =

    work page doi:10.1016/j.eswa.2024.126170 2025

  42. [43]

    Learning Dynamic Brain Network Representation Based on Graph Mamba Architecture , year=

    Wang, Jingjie and Lang, Jinwei and Yang, Li-Zhuang and Li, Hai , booktitle=. Learning Dynamic Brain Network Representation Based on Graph Mamba Architecture , year=

    work page

  43. [44]

    Spatial-Mamba: Effective Visual State Space Models via Structure-Aware State Fusion , volume =

    Xiao, Chaodong and Li, Minghan and ZHANG, zhengqiang and Meng, Deyu and Zhang, Lei , booktitle =. Spatial-Mamba: Effective Visual State Space Models via Structure-Aware State Fusion , volume =

    work page

  44. [45]

    Proceedings of the Computer Vision and Pattern Recognition Conference , pages=

    Mambairv2: Attentive state space restoration , author=. Proceedings of the Computer Vision and Pattern Recognition Conference , pages=

    work page

  45. [46]

    International Conference on Medical Image Computing and Computer-Assisted Intervention , pages=

    MNM: Multi-level Neuroimaging Meta-analysis with Hyperbolic Brain-Text Representations , author=. International Conference on Medical Image Computing and Computer-Assisted Intervention , pages=. 2025 , organization=

    work page 2025

  46. [47]

    IEEE Transactions on Consumer Electronics , year=

    Multi-scale graph transformer for brain disorder diagnosis , author=. IEEE Transactions on Consumer Electronics , year=

    work page

  47. [48]

    Graph Attention Networks , booktitle =

    Petar Velickovic and Guillem Cucurull and Arantxa Casanova and Adriana Romero and Pietro Li. Graph Attention Networks , booktitle =. 2018 , timestamp =

    work page 2018

  48. [49]

    Neural Multi-Objective Combinatorial Optimization via Graph-Image Multimodal Fusion , volume =

    Chen, Jinbiao and Wang, Jiahai and Cao, Zhiguang and Wu, Yaoxin , booktitle =. Neural Multi-Objective Combinatorial Optimization via Graph-Image Multimodal Fusion , volume =

    work page

  49. [50]

    Advances in neural information processing systems , volume=

    Attention is all you need , author=. Advances in neural information processing systems , volume=

    work page

  50. [51]

    Kingma and Jimmy Ba , editor =

    Diederik P. Kingma and Jimmy Ba , editor =. Adam:. 3rd International Conference on Learning Representations,. 2015 , url =

    work page 2015

  51. [52]

    Journal of neurophysiology , year=

    The organization of the human cerebral cortex estimated by intrinsic functional connectivity , author=. Journal of neurophysiology , year=

    work page

  52. [53]

    International Conference on Medical Image Computing and Computer-Assisted Intervention , pages=

    DHGFormer: Dynamic Hierarchical Graph Transformer for Disorder Brain Disease Diagnosis , author=. International Conference on Medical Image Computing and Computer-Assisted Intervention , pages=. 2025 , organization=

    work page 2025

  53. [54]

    and Funke, Michael and Mosher, John C

    Baker, Cole and Suárez-Méndez, Isabel and Smith, Grace and Marsh, Elisabeth B. and Funke, Michael and Mosher, John C. and Maestú, Fernando and Xu, Mengjia and Pantazis, Dimitrios , journal=. Hyperbolic Graph Embedding of MEG Brain Networks to Study Brain Alterations in Individuals With Subjective Cognitive Decline , year=

    work page

  54. [55]

    Transformers are

    Dao, Tri and Gu, Albert , booktitle =. Transformers are. 2024 , editor =

    work page 2024

  55. [56]

    Biological psychiatry , volume=

    Connectome-wide mega-analysis reveals robust patterns of atypical functional connectivity in autism , author=. Biological psychiatry , volume=. 2023 , publisher=

    work page 2023

  56. [57]

    Brain Connectivity , volume=

    Functional alterations associated with structural abnormalities in adults with high-functioning autism spectrum disorder , author=. Brain Connectivity , volume=. 2020 , publisher=

    work page 2020

  57. [58]

    Autism Research , volume=

    Specific functional connectivity patterns of middle temporal gyrus subregions in children and adults with autism spectrum disorder , author=. Autism Research , volume=. 2020 , publisher=

    work page 2020

  58. [59]

    Frontiers in neuroscience , volume=

    Altered functional connectivity in children with low-function autism spectrum disorders , author=. Frontiers in neuroscience , volume=. 2019 , publisher=

    work page 2019

  59. [60]

    JAMA psychiatry , volume=

    Large-scale network dysfunction in major depressive disorder: a meta-analysis of resting-state functional connectivity , author=. JAMA psychiatry , volume=

    work page

  60. [61]

    Proceedings of the National Academy of Sciences , volume=

    Resting-state functional MRI in depression unmasks increased connectivity between networks via the dorsal nexus , author=. Proceedings of the National Academy of Sciences , volume=. 2010 , publisher=

    work page 2010

  61. [62]

    Translational psychiatry , volume=

    Altered resting-state functional connectome in major depressive disorder: a mega-analysis from the PsyMRI consortium , author=. Translational psychiatry , volume=. 2021 , publisher=

    work page 2021

  62. [63]

    Biological psychiatry: cognitive neuroscience and neuroimaging , volume=

    Functional connectivity of the precuneus in unmedicated patients with depression , author=. Biological psychiatry: cognitive neuroscience and neuroimaging , volume=. 2018 , publisher=

    work page 2018

  63. [64]

    NeuroImage: Clinical , volume=

    Reduced functional connectivity of default mode network subsystems in depression: Meta-analytic evidence and relationship with trait rumination , author=. NeuroImage: Clinical , volume=. 2021 , publisher=

    work page 2021

  64. [65]

    Classification of Brain Disorders in rs-fMRI via Local-to-Global Graph Neural Networks , year=

    Zhang, Hao and Song, Ran and Wang, Liping and Zhang, Lin and Wang, Dawei and Wang, Cong and Zhang, Wei , journal=. Classification of Brain Disorders in rs-fMRI via Local-to-Global Graph Neural Networks , year=

    work page

  65. [66]

    Calhoun and Jingyu Liu , keywords =

    Bishal Thapaliya and Esra Akbas and Jiayu Chen and Ram Sapkota and Bhaskar Ray and Pranav Suresh and Vince D. Calhoun and Jingyu Liu , keywords =. Brain networks and intelligence: A graph neural network based approach to resting state fMRI data , journal =. 2025 , issn =

    work page 2025

  66. [67]

    MetaExplainer: Revisit domain generalization of functional connectome analyses from the perspective of explainability , journal =

    Xinmei Qiu and Yongheng Sun and Yilin Shi and Xujun Duan and Fan Wang and Jianhua Ma , keywords =. MetaExplainer: Revisit domain generalization of functional connectome analyses from the perspective of explainability , journal =. 2025 , issn =

    work page 2025

  67. [68]

    , title =

    Fortunato, S. , title =. Phys. Rep.-Rev. Sec. Phys. Lett. , volume =. 2010 , pages =

    work page 2010

  68. [69]

    Newman, M. E. J. and Girvan, M. , title =. Phys. Rev. E. , volume =. 2004 , pages =

    work page 2004

  69. [70]

    and Reinhardt, T

    Vehlow, C. and Reinhardt, T. and Weiskopf, D. , title =. IEEE Trans. Vis. Comput. Graph. , volume =. 2013 , pages =

    work page 2013

  70. [71]

    and Albert, R

    Raghavan, U. and Albert, R. and Kumara, S. , title =. Phys. Rev E. , volume =. 2007 , pages =

    work page 2007

  71. [72]

    2011 , pages =

    Robust network community detection using balanced propagation , journal =. 2011 , pages =

    work page 2011

  72. [73]

    and Li, S

    Lou, H. and Li, S. and Zhao, Y. , title =. Physica A. , volume =. 2013 , pages =

    work page 2013

  73. [74]

    and Newman, M

    Clauset, A. and Newman, M. E. J. and Moore, C. , title =. Phys. Rev. E. , volume =. 2004 , pages =

    work page 2004

  74. [75]

    Blondel, V. D. and Guillaume, J. L. and Lambiotte, R. and Lefebvre, E. , title =. J. Stat. Mech.-Theory Exp. , volume =. 2008 , pages =

    work page 2008

  75. [76]

    and Campari, R

    Sobolevsky, S. and Campari, R. , title =. Phys. Rev. E. , volume =. 2014 , pages =

    work page 2014

  76. [77]

    and Barthelemy, M

    Fortunato, S. and Barthelemy, M. , title =. Proc. Natl. Acad. Sci. U. S. A. , volume =. 2007 , pages =

    work page 2007

  77. [78]

    2011 , pages =

    Unfolding communities in large complex networks: Combining defensive and offensive label propagation for core extraction , journal =. 2011 , pages =

    work page 2011

  78. [79]

    and Li, J

    Wang, X. and Li, J. , title =. Physica A. , volume =. 2013 , pages =

    work page 2013

  79. [80]

    and Wang, X

    Li, J. and Wang, X. and Eustace, J. , title =. Physica A. , volume =. 2013 , pages =

    work page 2013

  80. [81]

    Fabio, D. R. and Fabio, D. and Carlo, P. , title =. Sci. Rep. , volume =. 2013 , pages =

    work page 2013

Showing first 80 references.