pith. sign in

arxiv: 2405.00577 · v2 · submitted 2024-05-01 · 💻 cs.LG · eess.SP· q-bio.NC

Discovering robust biomarkers of psychiatric disorders from resting-state functional MRI via graph neural networks: A systematic review

Yi Hao Chan , Deepank Girish , Sukrit Gupta , Jing Xia , Chockalingam Kasi , Yinan He , Conghao Wang , Jagath C. Rajapakse This is my paper

Pith reviewed 2026-05-24 01:07 UTC · model grok-4.3

classification 💻 cs.LG eess.SPq-bio.NC
keywords graph neural networksfMRI biomarkerspsychiatric disordersfeature attributionreproducibilitysystematic reviewresting-state fMRIdisorder classification
0
0 comments X

The pith

Graph neural network studies on fMRI biomarkers for psychiatric disorders produce inconsistent salient features across papers on the same condition.

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

This systematic review examines 65 studies that apply graph neural networks to resting-state fMRI data for classifying attention-deficit hyperactivity disorder, autism spectrum disorder, major depressive disorder, and schizophrenia. The authors observe that models often reach strong classification performance, yet the brain regions or connections flagged as important by feature attribution methods differ substantially from one study to the next for any given disorder. Only a small number of regions show repeated salience, and shared biomarkers across disorders remain rare. The paper concludes that current reliance on literature cross-referencing is too subjective and calls for objective evaluation metrics plus a structured prediction-attribution-evaluation process to improve biomarker robustness.

Core claim

In a review of 65 published studies that used graph neural networks on resting-state functional MRI to predict psychiatric disorders, salient features identified through feature attribution scores varied widely across studies examining the same disorder. Reproducibility was restricted to a limited set of regions, and few biomarkers appeared across multiple disorders. The authors therefore recommend shifting from subjective literature checks to standardized, objective metrics for assessing biomarker stability and introduce a prediction-attribution-evaluation framework to guide future work.

What carries the argument

Systematic comparison of feature attribution scores produced by graph neural networks across 65 independent studies on four psychiatric disorders.

If this is right

  • Classification accuracy is frequently high, yet the highlighted biomarkers remain inconsistent at the region level.
  • Only a small subset of regions demonstrates reproducible salience within each disorder.
  • Transdiagnostic biomarkers appear infrequently across the four conditions examined.
  • Current evaluation practices based on literature cross-referencing lack objectivity.
  • A prediction-attribution-evaluation framework is needed to establish more reliable standards.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • Differences in preprocessing pipelines or network construction choices may contribute to the observed feature variability beyond true biological differences.
  • Adopting objective metrics could reduce wasted effort on non-replicable findings and speed progress toward clinically useful biomarkers.
  • The same reproducibility challenges may affect other machine-learning approaches to neuroimaging beyond graph neural networks.

Load-bearing premise

The 65 reviewed studies form a representative sample and their feature attribution techniques reliably surface biologically meaningful signals rather than model-specific artifacts.

What would settle it

A controlled experiment that applies multiple distinct graph neural network architectures to identical fMRI datasets and finds the same top-ranked regions or edges across all models would contradict the reported high variability.

Figures

Figures reproduced from arXiv: 2405.00577 by Chockalingam Kasi, Conghao Wang, Deepank Girish, Jagath C. Rajapakse, Jing Xia, Sukrit Gupta, Yi Hao Chan, Yinan He.

Figure 1
Figure 1. Figure 1: Flowchart detailing the selection process of this review. 2. Modelling functional MRI datasets for disorder prediction Blood-oxygen-level-dependent signals captured in fMRI scans are fundamentally represented as time series data from individual voxels. Even at relatively low resolutions (e.g. 5mm), the number of voxels (> 10, 000) far outnumbers typical dataset sizes. Coupled with the issue of low signal-t… view at source ↗
Figure 2
Figure 2. Figure 2: Summary of the key components of typical GNNs used on fMRI datasets. and Betzel, 2020) (via much larger FC matrices, as all pairwise combinations of nodes are considered). Going beyond pairwise relationships, hypergraphs are higher-order graphs where each hyperedge represents the relationship between two or more nodes. In recent studies, hypergraphs based on connectomes are typically generated dynamically … view at source ↗
Figure 3
Figure 3. Figure 3: Our proposed taxonomy of state-of-the-art GNN models customised for fMRI datasets. mechanism, (iii) techniques used to train these GNNs [PITH_FULL_IMAGE:figures/full_fig_p010_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Taxonomy of attributors applicable to GNNs. Generally, post-hoc methods (typically ‘black box’) have a larger range of algorithms than self-interpretable (typically ‘grey box’) ones. Methods that have not been explored in fMRI studies are highlighted in yellow. be extracted by identifying features that have the largest coefficients assigned to them by the model fitting process. On the other end of the spec… view at source ↗
Figure 5
Figure 5. Figure 5: A subset of the Co-12 properties (Nauta et al., 2023) used to evaluate feature attribution scores produced by attributors, deemed to be relevant for biomarker discovery from fMRI datasets. evaluating correctness involve deletion: changes in model outputs are computed for each feature subset (in the simplest case, independently removing each feature in the node vector) and then correlated with the importanc… view at source ↗
Figure 6
Figure 6. Figure 6: An illustration showing where the three key stages occur in a typical pipeline that uses a post-hoc attributor. For intrinsically interpretable models, the attributor would become part of the predictor (e.g. pooling layers can be used for producing explanations, but they form part of the GNN architecture). In order to discover robust biomarkers of psychiatric disorders via machine learning techniques, it i… view at source ↗
Figure 7
Figure 7. Figure 7: Plot of model accuracy against dataset size. It is evident that a majority of the studies have dataset sizes below 1000 and classification accuracies generally decrease as the dataset size increases. In view of this, we reiterate the need for proper benchmarking. Although several benchmarking studies have been done using GNNs on fMRI datasets, they were focused on baseline GNNs and not state-of-the-art mod… view at source ↗
Figure 8
Figure 8. Figure 8: Choice of GNN used in the papers reviewed. Some studies included more than one type of GNN. ‘New’ refers to architectures that significantly deviate from baseline GNNs (e.g. customised message passing techniques). Colour scheme represents various classes of GNNs - Light red: GCN (spatial/spectral), Shades of yellow: Spatial GNN, Blue: Spectral GNN, Shades of orange: Others) [PITH_FULL_IMAGE:figures/full_f… view at source ↗
Figure 9
Figure 9. Figure 9: Distribution of attributors covered in our review. Some studies used more than one type of attributor. Colour scheme represents various categories of attributors, following our proposed taxonomy - Shades of red: Self-interpretable, Orange: Others, Shades of yellow: Post-hoc, Shades of blue: ‘Traditional’ approaches [PITH_FULL_IMAGE:figures/full_fig_p046_9.png] view at source ↗
read the original abstract

Graph neural networks (GNN) have emerged as a popular tool for modelling functional magnetic resonance imaging (fMRI) datasets. Many recent studies have reported significant improvements in disorder classification performance via more sophisticated GNN designs and highlighted salient features that could be potential biomarkers of the disorder. However, existing methods of evaluating their robustness are often limited to cross-referencing with existing literature, which is a subjective and inconsistent process. In this review, we provide an overview of how GNN and model explainability techniques (specifically, feature attributors) have been applied to fMRI datasets for disorder prediction tasks, with an emphasis on evaluating the robustness of potential biomarkers produced for psychiatric disorders. Then, 65 studies using GNNs that reported potential fMRI biomarkers for psychiatric disorders (attention-deficit hyperactivity disorder, autism spectrum disorder, major depressive disorder, schizophrenia) published before 9 October 2024 were identified from 2 online databases (Scopus, PubMed). We found that while most studies have performant models, salient features highlighted in these studies (as determined by feature attribution scores) vary greatly across studies on the same disorder. Reproducibility of biomarkers is only limited to a small subset at the level of regions and few transdiagnostic biomarkers were identified. To address these issues, we suggest establishing new standards that are based on objective evaluation metrics to determine the robustness of these potential biomarkers. We further highlight gaps in the existing literature and put together a prediction-attribution-evaluation framework that could set the foundations for future research on discovering robust biomarkers of psychiatric disorders via GNNs.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 2 minor

Summary. This systematic review surveys 65 studies applying graph neural networks (GNNs) to resting-state fMRI for classifying four psychiatric disorders (ADHD, ASD, MDD, schizophrenia). The central claims are that models are generally performant yet the salient regions identified by feature attribution methods vary substantially across studies on the same disorder, that region-level reproducibility is limited to small subsets, and that few transdiagnostic biomarkers emerge; the authors therefore call for objective evaluation standards and outline a prediction-attribution-evaluation framework.

Significance. If the cross-study synthesis is methodologically sound, the work usefully documents limited biomarker reproducibility in the GNN-fMRI literature and supplies a concrete framework that could improve future studies. The review also correctly notes that reliance on subjective literature cross-referencing is inadequate and that objective metrics are needed.

major comments (2)
  1. [Methods / Results synthesis] The central claim that attribution-derived salient features 'vary greatly' and show 'limited reproducibility' rests on an unstandardized aggregation across 65 studies that differ in atlas (AAL, Schaefer, etc.), explainer (GNNExplainer, Grad-CAM, attention), and reporting format. No section of the manuscript describes an explicit harmonization protocol (anatomical label mapping, overlap metric, sensitivity analysis to k/threshold, or inter-atlas normalization), so measured variation may conflate methodological mismatch with biological non-reproducibility.
  2. [Abstract / Methods] The abstract and review summary state that 65 studies were identified from Scopus and PubMed before 9 October 2024, yet supply no search string, inclusion/exclusion criteria, or inter-rater reliability statistics for feature extraction. Without these details the representativeness of the sample and the reliability of the extracted biomarker lists cannot be assessed.
minor comments (2)
  1. [Discussion] The proposed prediction-attribution-evaluation framework is only sketched at a high level; a concrete workflow diagram or pseudocode would clarify how the suggested objective metrics would be applied.
  2. [Results] Table or supplementary material listing the 65 studies with their atlases, explainers, and reported top regions would allow readers to verify the variability claim directly.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the detailed and constructive report. The two major comments highlight important gaps in methodological transparency that we will address in revision. Below we respond point-by-point.

read point-by-point responses

  1. Referee: [Methods / Results synthesis] The central claim that attribution-derived salient features 'vary greatly' and show 'limited reproducibility' rests on an unstandardized aggregation across 65 studies that differ in atlas (AAL, Schaefer, etc.), explainer (GNNExplainer, Grad-CAM, attention), and reporting format. No section of the manuscript describes an explicit harmonization protocol (anatomical label mapping, overlap metric, sensitivity analysis to k/threshold, or inter-atlas normalization), so measured variation may conflate methodological mismatch with biological non-reproducibility.

    Authors: We agree that the absence of an explicit harmonization protocol is a limitation. While our synthesis already notes atlas and explainer heterogeneity as a source of variation, we did not document the precise mapping rules or sensitivity checks applied during aggregation. In the revised manuscript we will add a new subsection (Methods, 'Biomarker harmonization and sensitivity analysis') that (i) describes the anatomical label mapping procedure between common atlases, (ii) specifies the overlap metric (Jaccard index at region level) and threshold (top-k = 10 regions), and (iii) reports a sensitivity analysis repeating the reproducibility counts under alternative k values and atlas normalizations. This will allow readers to assess how much of the observed non-reproducibility is attributable to methodological mismatch versus biological signal. revision: yes

  2. Referee: [Abstract / Methods] The abstract and review summary state that 65 studies were identified from Scopus and PubMed before 9 October 2024, yet supply no search string, inclusion/exclusion criteria, or inter-rater reliability statistics for feature extraction. Without these details the representativeness of the sample and the reliability of the extracted biomarker lists cannot be assessed.

    Authors: The referee is correct; these details are missing from the current text. The abstract and Methods section mention only the two databases and cutoff date. In revision we will expand the Methods section with: (a) the exact search strings used in Scopus and PubMed, (b) a PRISMA-style flow diagram and explicit inclusion/exclusion criteria (e.g., must apply a GNN to rs-fMRI, report feature-attribution results for one of the four disorders, peer-reviewed, etc.), and (c) a statement on how biomarker lists were extracted (single reviewer with spot-checks by a second author; no formal inter-rater statistic was computed). These additions will be placed before the results on biomarker reproducibility. revision: yes

Circularity Check

0 steps flagged

Systematic review exhibits no circularity; synthesis grounded in external literature

full rationale

This is a systematic review paper that identifies and summarizes 65 external studies without performing any model fitting, deriving predictions from its own parameters, or invoking self-citation chains as load-bearing premises. The reported observations on biomarker variation are direct aggregations of findings from the cited literature, with no equations, ansatzes, or uniqueness theorems that reduce by construction to the paper's own inputs. The central claims remain independent of any internal self-referential structure.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The review rests on standard assumptions of systematic review methodology (comprehensive search, unbiased selection) and on the premise that feature attribution scores in the source papers reflect stable biological signals. No free parameters or invented entities are introduced.

axioms (1)
  • domain assumption Feature attribution methods applied to GNNs on fMRI data produce comparable and biologically interpretable salience maps across different studies and model architectures.
    Invoked when the review treats variation in reported salient features as evidence of non-robustness rather than method-specific artifacts.

pith-pipeline@v0.9.0 · 5855 in / 1322 out tokens · 18952 ms · 2026-05-24T01:07:35.854252+00:00 · methodology

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

208 extracted references · 208 canonical work pages · 1 internal anchor

  1. [1]

    , author Moeller, S.J

    author Abi-Dargham, A. , author Moeller, S.J. , author Ali, F. , author DeLorenzo, C. , author Domschke, K. , author Horga, G. , author Jutla, A. , author Kotov, R. , author Paulus, M.P. , author Rubio, J.M. , et al., year 2023 . title Candidate biomarkers in psychiatric disorders: state of the field . journal World Psychiatry volume 22 , pages 236--262

    work page 2023

  2. [2]

    , author Queen, O

    author Agarwal, C. , author Queen, O. , author Lakkaraju, H. , author Zitnik, M. , year 2023 . title Evaluating explainability for graph neural networks . journal Scientific Data volume 10 , pages 144

    work page 2023

  3. [3]

    , author Abuhmed, T

    author Ali, S. , author Abuhmed, T. , author El-Sappagh, S. , author Muhammad, K. , author Alonso-Moral, J.M. , author Confalonieri, R. , author Guidotti, R. , author Del Ser, J. , author D \' az-Rodr \' guez, N. , author Herrera, F. , year 2023 . title Explainable artificial intelligence (xai): What we know and what is left to attain trustworthy artifici...

    work page 2023

  4. [4]

    , author Robbins, T.W

    author Aron, A.R. , author Robbins, T.W. , author Poldrack, R.A. , year 2004 . title Inhibition and the right inferior frontal cortex . journal Trends in cognitive sciences volume 8 , pages 170--177

    work page 2004

  5. [5]

    , author Ferner, R.E

    author Aronson, J.K. , author Ferner, R.E. , year 2017 . title Biomarkers—a general review . journal Current protocols in pharmacology volume 76 , pages 9--23

    work page 2017

  6. [6]

    , author Olij, R

    author Arya, D. , author Olij, R. , author Gupta, D.K. , author El Gazzar, A. , author Wingen, G. , author Worring, M. , author Thomas, R.M. , year 2020 . title Fusing structural and functional mris using graph convolutional networks for autism classification , in: booktitle Medical imaging with deep learning , organization PMLR . pp. pages 44--61

    work page 2020

  7. [7]

    , year 2024

    author Atanasova, P. , year 2024 . title A diagnostic study of explainability techniques for text classification , in: booktitle Accountable and Explainable Methods for Complex Reasoning over Text . publisher Springer , pp. pages 155--187

    work page 2024

  8. [8]

    , author Liang, J

    author Bai, B. , author Liang, J. , author Zhang, G. , author Li, H. , author Bai, K. , author Wang, F. , year 2021 a. title Why attentions may not be interpretable? , in: booktitle Proceedings of the 27th ACM SIGKDD conference on knowledge discovery & data mining , pp. pages 25--34

    work page 2021

  9. [9]

    , author Zhang, F

    author Bai, S. , author Zhang, F. , author Torr, P.H. , year 2021 b. title Hypergraph convolution and hypergraph attention . journal Pattern Recognition volume 110 , pages 107637

    work page 2021

  10. [10]

    , author Thompson, P.M

    author Bayer, J.M. , author Thompson, P.M. , author Ching, C.R. , author Liu, M. , author Chen, A. , author Panzenhagen, A.C. , author Jahanshad, N. , author Marquand, A. , author Schmaal, L. , author S \"a mann, P.G. , year 2022 . title Site effects how-to and when: An overview of retrospective techniques to accommodate site effects in multi-site neuroim...

    work page 2022

  11. [11]

    , author Mahjoub, M.A

    author Bessadok, A. , author Mahjoub, M.A. , author Rekik, I. , year 2022 . title Graph neural networks in network neuroscience . journal IEEE Transactions on Pattern Analysis and Machine Intelligence volume 45 , pages 5833--5848

    work page 2022

  12. [12]

    , author Xia, N

    author Bian, C. , author Xia, N. , author Xie, A. , author Cong, S. , author Dong, Q. , year 2024 . title Adversarially trained persistent homology based graph convolutional network for disease identification using brain connectivity . journal IEEE transactions on medical imaging volume 43 , pages 503--516

    work page 2024

  13. [13]

    , author Uithol, S

    author Bielczyk, N.Z. , author Uithol, S. , author van Mourik, T. , author Anderson, P. , author Glennon, J.C. , author Buitelaar, J.K. , year 2019 . title Disentangling causal webs in the brain using functional magnetic resonance imaging: A review of current approaches . journal Network Neuroscience volume 3 , pages 237--273

    work page 2019

  14. [14]

    , author Mueller, T.T

    author Bintsi, K.M. , author Mueller, T.T. , author Starck, S. , author Baltatzis, V. , author Hammers, A. , author Rueckert, D. , year 2023 . title A comparative study of population-graph construction methods and graph neural networks for brain age regression , in: booktitle International Conference on Medical Image Computing and Computer-Assisted Interv...

    work page 2023

  15. [15]

    , author Maggioni, E

    author Bondi, E. , author Maggioni, E. , author Brambilla, P. , author Delvecchio, G. , year 2023 . title A systematic review on the potential use of machine learning to classify major depressive disorder from healthy controls using resting state fmri measures . journal Neuroscience & Biobehavioral Reviews volume 144 , pages 104972

    work page 2023

  16. [16]

    , author Holzmeister, F

    author Botvinik-Nezer, R. , author Holzmeister, F. , author Camerer, C.F. , author Dreber, A. , author Huber, J. , author Johannesson, M. , author Kirchler, M. , author Iwanir, R. , author Mumford, J.A. , author Adcock, R.A. , et al., year 2020 . title Variability in the analysis of a single neuroimaging dataset by many teams . journal Nature volume 582 ,...

    work page 2020

  17. [17]

    , author Liu, J

    author Calhoun, V.D. , author Liu, J. , author Adal , T. , year 2009 . title A review of group ica for fmri data and ica for joint inference of imaging, genetic, and erp data . journal Neuroimage volume 45 , pages S163--S172

    work page 2009

  18. [18]

    , year 2018

    author Califf, R.M. , year 2018 . title Biomarker definitions and their applications . journal Experimental biology and medicine volume 243 , pages 213--221

    work page 2018

  19. [19]

    , author Chen, D

    author Canario, E. , author Chen, D. , author Biswal, B. , year 2021 . title A review of resting-state fmri and its use to examine psychiatric disorders . journal Psychoradiology volume 1 , pages 42--53

    work page 2021

  20. [20]

    , author M \"u ller, V.I

    author Cash, R.F. , author M \"u ller, V.I. , author Fitzgerald, P.B. , author Eickhoff, S.B. , author Zalesky, A. , year 2023 . title Altered brain activity in unipolar depression unveiled using connectomics . journal Nature Mental Health volume 1 , pages 174--185

    work page 2023

  21. [21]

    , author Yew, W.C

    author Chan, Y.H. , author Yew, W.C. , author Chew, Q.H. , author Sim, K. , author Rajapakse, J.C. , year 2023 . title Elucidating salient site-specific functional connectivity features and site-invariant biomarkers in schizophrenia via deep neural networks . journal Scientific Reports volume 13 , pages 21047

    work page 2023

  22. [22]

    , author Yew, W.C

    author Chan, Y.H. , author Yew, W.C. , author Rajapakse, J.C. , year 2022 . title Semi-supervised learning with data harmonisation for biomarker discovery from resting state fmri , in: booktitle International Conference on Medical Image Computing and Computer-Assisted Intervention , organization Springer . pp. pages 441--451

    work page 2022

  23. [23]

    , author Patil, K.R

    author Chen, J. , author Patil, K.R. , author Yeo, B.T. , author Eickhoff, S.B. , year 2023 a. title Leveraging machine learning for gaining neurobiological and nosological insights in psychiatric research . journal Biological psychiatry volume 93 , pages 18--28

    work page 2023

  24. [24]

    , author Ke, P

    author Chen, X. , author Ke, P. , author Huang, Y. , author Zhou, J. , author Li, H. , author Peng, R. , author Huang, J. , author Liang, L. , author Ma, G. , author Li, X. , et al., year 2023 b. title Discriminative analysis of schizophrenia patients using graph convolutional networks: A combined multimodal mri and connectomics analysis . journal Frontie...

    work page 2023

  25. [25]

    , author Liu, A

    author Chen, Y. , author Liu, A. , author Fu, X. , author Wen, J. , author Chen, X. , year 2022 . title An invertible dynamic graph convolutional network for multi-center asd classification . journal Frontiers in Neuroscience volume 15 , pages 828512

    work page 2022

  26. [26]

    , author Yan, J

    author Chen, Y. , author Yan, J. , author Jiang, M. , author Zhang, T. , author Zhao, Z. , author Zhao, W. , author Zheng, J. , author Yao, D. , author Zhang, R. , author Kendrick, K.M. , author Jiang, X. , year 2024 . title Adversarial learning based node-edge graph attention networks for autism spectrum disorder identification . journal IEEE Transaction...

    work page doi:10.1109/tnnls.2022.3154755 2024

  27. [27]

    , author Yan, J

    author Chen, Y. , author Yan, J. , author Jiang, M. , author Zhao, Z. , author Zhao, W. , author Zhang, R. , author Kendrick, K.M. , author Jiang, X. , year 2021 . title Attention-based node-edge graph convolutional networks for identification of autism spectrum disorder using multi-modal mri data , in: booktitle Pattern Recognition and Computer Vision: 4...

    work page 2021

  28. [28]

    , author Payoux, P

    author Chollet, F. , author Payoux, P. , year 2022 . title Functional imaging for neurodegenerative diseases . journal La Presse M \'e dicale volume 51 , pages 104121

    work page 2022

  29. [29]

    , author Ren, H

    author Chu, Y. , author Ren, H. , author Qiao, L. , author Liu, M. , year 2022 . title Resting-state functional mri adaptation with attention graph convolution network for brain disorder identification . journal Brain Sciences volume 12 , pages 1413

    work page 2022

  30. [30]

    , author Kazi, A

    author Cosmo, L. , author Kazi, A. , author Ahmadi, S.A. , author Navab, N. , author Bronstein, M. , year 2020 . title Latent-graph learning for disease prediction , in: booktitle Medical Image Computing and Computer Assisted Intervention--MICCAI 2020: 23rd International Conference, Lima, Peru, October 4--8, 2020, Proceedings, Part II 23 , organization Sp...

    work page 2020

  31. [31]

    , author Thompson-Schill, S.L

    author Coutanche, M.N. , author Thompson-Schill, S.L. , author Schultz, R.T. , year 2011 . title Multi-voxel pattern analysis of fmri data predicts clinical symptom severity . journal Neuroimage volume 57 , pages 113--123

    work page 2011

  32. [32]

    , author Dai, W

    author Cui, H. , author Dai, W. , author Zhu, Y. , author Kan, X. , author Gu, A.A.C. , author Lukemire, J. , author Zhan, L. , author He, L. , author Guo, Y. , author Yang, C. , year 2022 . title Braingb: A benchmark for brain network analysis with graph neural networks . journal IEEE transactions on medical imaging volume 42 , pages 493--506

    work page 2022

  33. [33]

    , author Du, J

    author Cui, W. , author Du, J. , author Sun, M. , author Zhu, S. , author Zhao, S. , author Peng, Z. , author Tan, L. , author Li, Y. , year 2023 . title Dynamic multi-site graph convolutional network for autism spectrum disorder identification . journal Computers in Biology and Medicine volume 157 , pages 106749

    work page 2023

  34. [34]

    , author Rahim, M

    author Dadi, K. , author Rahim, M. , author Abraham, A. , author Chyzhyk, D. , author Milham, M. , author Thirion, B. , author Varoquaux, G. , author Initiative, A.D.N. , et al., year 2019 . title Benchmarking functional connectome-based predictive models for resting-state fmri . journal NeuroImage volume 192 , pages 115--134

    work page 2019

  35. [35]

    , author Lu, D

    author Dai, P. , author Lu, D. , author Shi, Y. , author Zhou, Y. , author Xiong, T. , author Zhou, X. , author Chen, Z. , author Zou, B. , author Tang, H. , author Huang, Z. , et al., year 2023 . title Classification of recurrent major depressive disorder using a new time series feature extraction method through multisite rs-fmri data . journal Journal o...

    work page 2023

  36. [36]

    , author Bresson, X

    author Defferrard, M. , author Bresson, X. , author Vandergheynst, P. , year 2016 . title Convolutional neural networks on graphs with fast localized spectral filtering . journal Advances in neural information processing systems volume 29

    work page 2016

  37. [37]

    , author Chamberlain, S.R

    author Del Campo, N. , author Chamberlain, S.R. , author Sahakian, B.J. , author Robbins, T.W. , year 2011 . title The roles of dopamine and noradrenaline in the pathophysiology and treatment of attention-deficit/hyperactivity disorder . journal Biological psychiatry volume 69 , pages e145--e157

    work page 2011

  38. [38]

    , author S \'e gonne, F

    author Desikan, R.S. , author S \'e gonne, F. , author Fischl, B. , author Quinn, B.T. , author Dickerson, B.C. , author Blacker, D. , author Buckner, R.L. , author Dale, A.M. , author Maguire, R.P. , author Hyman, B.T. , et al., year 2006 . title An automated labeling system for subdividing the human cerebral cortex on mri scans into gyral based regions ...

    work page 2006

  39. [39]

    , author Fu, Z

    author Du, Y. , author Fu, Z. , author Calhoun, V.D. , year 2018 . title Classification and prediction of brain disorders using functional connectivity: promising but challenging . journal Frontiers in neuroscience volume 12 , pages 525

    work page 2018

  40. [40]

    , author Iraji, A

    author Duda, M. , author Iraji, A. , author Ford, J.M. , author Lim, K.O. , author Mathalon, D.H. , author Mueller, B.A. , author Potkin, S.G. , author Preda, A. , author Van Erp, T.G. , author Calhoun, V.D. , year 2023 . title Reliability and clinical utility of spatially constrained estimates of intrinsic functional networks from very short fmri scans ....

    work page 2023

  41. [41]

    , author Thomas, R

    author ElGazzar, A. , author Thomas, R. , author Van Wingen, G. , year 2022 . title Benchmarking graph neural networks for fmri analysis . journal arXiv preprint arXiv:2211.08927

    work page arXiv 2022

  42. [42]

    , author Byrge, L

    author Esfahlani, F.Z. , author Byrge, L. , author Tanner, J. , author Sporns, O. , author Kennedy, D.P. , author Betzel, R.F. , year 2022 . title Edge-centric analysis of time-varying functional brain networks with applications in autism spectrum disorder . journal NeuroImage volume 263 , pages 119591

    work page 2022

  43. [43]

    , author Markiewicz, C.J

    author Esteban, O. , author Markiewicz, C.J. , author Blair, R.W. , author Moodie, C.A. , author Isik, A.I. , author Erramuzpe, A. , author Kent, J.D. , author Goncalves, M. , author DuPre, E. , author Snyder, M. , et al., year 2019 . title fmriprep: a robust preprocessing pipeline for functional mri . journal Nature methods volume 16 , pages 111--116

    work page 2019

  44. [44]

    , author Su, J

    author Fan, Z. , author Su, J. , author Xu, J. , author Hu, D. , author Zeng, L.L. , year 2023 . title Dgst-former: Dynamic graph with spatio-temporal transpose-transformer for brain fmri data , in: booktitle 2023 7th Asian Conference on Artificial Intelligence Technology (ACAIT) , organization IEEE . pp. pages 530--535

    work page 2023

  45. [45]

    , author Zhang, D.f

    author Fang, J. , author Zhang, D.f. , author Xie, K. , author Xu, L. , author Bi, X.a. , year 2024 . title Bilinear perceptual fusion algorithm based on brain functional and structural data for asd diagnosis and regions of interest identification . journal Interdisciplinary Sciences: Computational Life Sciences , pages 1--15

    work page 2024

  46. [46]

    , author Wang, M

    author Fang, Y. , author Wang, M. , author Potter, G.G. , author Liu, M. , year 2023 . title Unsupervised cross-domain functional mri adaptation for automated major depressive disorder identification . journal Medical image analysis volume 84 , pages 102707

    work page 2023

  47. [47]

    , author Esfahlani, F.Z

    author Faskowitz, J. , author Esfahlani, F.Z. , author Jo, Y. , author Sporns, O. , author Betzel, R.F. , year 2020 . title Edge-centric functional network representations of human cerebral cortex reveal overlapping system-level architecture . journal Nature neuroscience volume 23 , pages 1644--1654

    work page 2020

  48. [48]

    , author You, C

    author Feng, A. , author You, C. , author Wang, S. , author Tassiulas, L. , year 2022 . title Kergnns: Interpretable graph neural networks with graph kernels , in: booktitle Proceedings of the AAAI conference on artificial intelligence , pp. pages 6614--6622

    work page 2022

  49. [49]

    , author Spinelli, E.G

    author Filippi, M. , author Spinelli, E.G. , author Cividini, C. , author Ghirelli, A. , author Basaia, S. , author Agosta, F. , year 2023 . title The human functional connectome in neurodegenerative diseases: relationship to pathology and clinical progression . journal Expert Review of Neurotherapeutics volume 23 , pages 59--73

    work page 2023

  50. [50]

    , author Carbone, E.A

    author de Filippis, R. , author Carbone, E.A. , author Gaetano, R. , author Bruni, A. , author Pugliese, V. , author Segura-Garcia, C. , author De Fazio, P. , year 2019 . title Machine learning techniques in a structural and functional mri diagnostic approach in schizophrenia: a systematic review . journal Neuropsychiatric disease and treatment , pages 1605--1627

    work page 2019

  51. [51]

    , author Laird, A.R

    author Fox, P.T. , author Laird, A.R. , author Fox, S.P. , author Fox, P.M. , author Uecker, A.M. , author Crank, M. , author Koenig, S.F. , author Lancaster, J.L. , year 2005 . title Brainmap taxonomy of experimental design: description and evaluation . journal Human brain mapping volume 25 , pages 185--198

    work page 2005

  52. [52]

    , author El-Gazzar, A

    author Gallo, S. , author El-Gazzar, A. , author Zhutovsky, P. , author Thomas, R.M. , author Javaheripour, N. , author Li, M. , author Bartova, L. , author Bathula, D. , author Dannlowski, U. , author Davey, C. , et al., year 2023 . title Functional connectivity signatures of major depressive disorder: machine learning analysis of two multicenter neuroim...

    work page 2023

  53. [53]

    , year 2023

    author Ghosal, S. , year 2023 . title Interpretable Machine Learning and Deep Learning Frameworks for Predictive Analytics and Biomarker Discovery from Multimodal Imaging Genetics Data . Ph.D. thesis. Johns Hopkins University

    work page 2023

  54. [54]

    , author Chan, Y.H

    author Girish, D. , author Chan, Y.H. , author Gupta, S. , author Xia, J. , author Rajapakse, J. , year 2024 . title Robustness of explainable ai algorithms for disease biomarker discovery from functional connectivity datasets , in: booktitle IEEE-EMBS International Conference on Biomedical and Health Informatics

    work page 2024

  55. [55]

    , author Auer, T

    author Gorgolewski, K.J. , author Auer, T. , author Calhoun, V.D. , author Craddock, R.C. , author Das, S. , author Duff, E.P. , author Flandin, G. , author Ghosh, S.S. , author Glatard, T. , author Halchenko, Y.O. , et al., year 2016 . title The brain imaging data structure, a format for organizing and describing outputs of neuroimaging experiments . jou...

    work page 2016

  56. [56]

    , author Peng, S

    author Gu, Y. , author Peng, S. , author Li, Y. , author Gao, L. , author Dong, Y. , year 2024 . title A novel population graph neural network based on functional connectivity for mental disorders detection , in: booktitle Pacific-Asia Conference on Knowledge Discovery and Data Mining , organization Springer . pp. pages 221--233

    work page 2024

  57. [57]

    , author Peng, S

    author Gu, Y. , author Peng, S. , author Li, Y. , author Gao, L. , author Dong, Y. , year 2025 . title Fc-hgnn: A heterogeneous graph neural network based on brain functional connectivity for mental disorder identification . journal Information Fusion volume 113 , pages 102619

    work page 2025

  58. [58]

    , author Chan, Y.H

    author Gupta, S. , author Chan, Y.H. , author Rajapakse, J.C. , year 2021 . title Obtaining leaner deep neural networks for decoding brain functional connectome in a single shot . journal Neurocomputing volume 453 , pages 326--336

    work page 2021

  59. [59]

    , author Lim, M

    author Gupta, S. , author Lim, M. , author Rajapakse, J.C. , year 2022 . title Decoding task specific and task general functional architectures of the brain . journal Human Brain Mapping volume 43 , pages 2801--2816

    work page 2022

  60. [60]

    , author Stern, Y

    author Habeck, C. , author Stern, Y. , year 2010 . title Multivariate data analysis for neuroimaging data: overview and application to alzheimer’s disease . journal Cell biochemistry and biophysics volume 58 , pages 53--67

    work page 2010

  61. [61]

    , author Wei, Z

    author He, M. , author Wei, Z. , author Wen, J.R. , year 2022 . title Convolutional neural networks on graphs with chebyshev approximation, revisited . journal Advances in neural information processing systems volume 35 , pages 7264--7276

    work page 2022

  62. [62]

    , author Chan, Y.H

    author He, Y. , author Chan, Y.H. , author Rajapakse, J.C. , year 2023 . title Predicting gender from structural and functional connectomes via brain and population graph neural networks . journal bioRxiv , pages 2023--11

    work page 2023

  63. [63]

    , author Sporns, O

    author van den Heuvel, M.P. , author Sporns, O. , year 2019 . title A cross-disorder connectome landscape of brain dysconnectivity . journal Nature reviews neuroscience volume 20 , pages 435--446

    work page 2019

  64. [64]

    , author Xu, T

    author Hong, S.J. , author Xu, T. , author Nikolaidis, A. , author Smallwood, J. , author Margulies, D.S. , author Bernhardt, B. , author Vogelstein, J. , author Milham, M.P. , year 2020 . title Toward a connectivity gradient-based framework for reproducible biomarker discovery . journal NeuroImage volume 223 , pages 117322

    work page 2020

  65. [65]

    , author Chen, A.A

    author Hu, F. , author Chen, A.A. , author Horng, H. , author Bashyam, V. , author Davatzikos, C. , author Alexander-Bloch, A. , author Li, M. , author Shou, H. , author Satterthwaite, T.D. , author Yu, M. , et al., year 2023 . title Image harmonization: A review of statistical and deep learning methods for removing batch effects and evaluation metrics fo...

    work page 2023

  66. [66]

    , author Cao, L

    author Hu, J. , author Cao, L. , author Li, T. , author Dong, S. , author Li, P. , year 2021 . title Gat-li: a graph attention network based learning and interpreting method for functional brain network classification . journal BMC bioinformatics volume 22 , pages 1--20

    work page 2021

  67. [67]

    , author Yamada, M

    author Huang, Q. , author Yamada, M. , author Tian, Y. , author Singh, D. , author Chang, Y. , year 2022 . title Graphlime: Local interpretable model explanations for graph neural networks . journal IEEE Transactions on Knowledge and Data Engineering volume 35 , pages 6968--6972

    work page 2022

  68. [68]

    , year 2010

    author Jablensky, A. , year 2010 . title The diagnostic concept of schizophrenia: its history, evolution, and future prospects . journal Dialogues in clinical neuroscience volume 12 , pages 271--287

    work page 2010

  69. [69]

    , author Jain, K.K

    author Jain, K.K. , author Jain, K.K. , year 2010 . title The handbook of biomarkers . volume volume 6 . publisher Springer

    work page 2010

  70. [70]

    , author Wallace, B.C

    author Jain, S. , author Wallace, B.C. , year 2019 . title Attention is not explanation , in: booktitle Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 1 (Long and Short Papers) , pp. pages 3543--3556

    work page 2019

  71. [71]

    , author Ren, Y

    author Ji, J. , author Ren, Y. , author Lei, M. , year 2022 . title Fc--hat: hypergraph attention network for functional brain network classification . journal Information Sciences volume 608 , pages 1301--1316

    work page 2022

  72. [72]

    , author Zhao, N

    author Jia, X.Z. , author Zhao, N. , author Barton, B. , author Burciu, R. , author Carri \`e re, N. , author Cerasa, A. , author Chen, B.Y. , author Chen, J. , author Coombes, S. , author Defebvre, L. , et al., year 2018 . title Small effect size leads to reproducibility failure in resting-state fmri studies . journal BioRxiv , pages 285171

    work page 2018

  73. [73]

    , author Cao, P

    author Jiang, H. , author Cao, P. , author Xu, M. , author Yang, J. , author Zaiane, O. , year 2020 . title Hi-gcn: A hierarchical graph convolution network for graph embedding learning of brain network and brain disorders prediction . journal Computers in Biology and Medicine volume 127 , pages 104096

    work page 2020

  74. [74]

    , author Na, K.S

    author Jun, E. , author Na, K.S. , author Kang, W. , author Lee, J. , author Suk, H.I. , author Ham, B.J. , year 2020 . title Identifying resting-state effective connectivity abnormalities in drug-na \" ve major depressive disorder diagnosis via graph convolutional networks . journal Human Brain Mapping volume 41 , pages 4997--5014

    work page 2020

  75. [75]

    , author Keefe, R.S

    author Kahn, R.S. , author Keefe, R.S. , year 2013 . title Schizophrenia is a cognitive illness: time for a change in focus . journal JAMA psychiatry volume 70 , pages 1107--1112

    work page 2013

  76. [76]

    , et al., year 2015

    author Kahn, R.S. , et al., year 2015 . title Schizophrenia . journal Nature Reviews Disease Primers volume 1 , pages 15067 . https://doi.org/10.1038/nrdp.2015.67, :10.1038/nrdp.2015.67

    work page doi:10.1038/nrdp.2015.67 2015

  77. [77]

    A survey on explain- ability of graph neural networks

    author Kakkad, J. , author Jannu, J. , author Sharma, K. , author Aggarwal, C. , author Medya, S. , year 2023 . title A survey on explainability of graph neural networks . journal arXiv preprint arXiv:2306.01958

    work page arXiv 2023

  78. [78]

    , author Venugopalan, S

    author Kapishnikov, A. , author Venugopalan, S. , author Avci, B. , author Wedin, B. , author Terry, M. , author Bolukbasi, T. , year 2021 . title Guided integrated gradients: An adaptive path method for removing noise , in: booktitle Proceedings of the IEEE/CVF conference on computer vision and pattern recognition , pp. pages 5050--5058

    work page 2021

  79. [79]

    , author Brown, C.J

    author Kawahara, J. , author Brown, C.J. , author Miller, S.P. , author Booth, B.G. , author Chau, V. , author Grunau, R.E. , author Zwicker, J.G. , author Hamarneh, G. , year 2017 . title Brainnetcnn: Convolutional neural networks for brain networks; towards predicting neurodevelopment . journal NeuroImage volume 146 , pages 1038--1049

    work page 2017

  80. [80]

    , author Welling, M

    author Kipf, T.N. , author Welling, M. , year 2017 . title Semi-supervised classification with graph convolutional networks , in: booktitle International Conference on Learning Representations . https://openreview.net/forum?id=SJU4ayYgl

    work page 2017

Showing first 80 references.