{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,9,23]],"date-time":"2026-09-23T07:01:49Z","timestamp":1790146909772,"version":"4.0.1"},"reference-count":53,"publisher":"Frontiers Media SA","license":[{"start":{"date-parts":[[2026,9,23]],"date-time":"2026-09-23T00:00:00Z","timestamp":1790121600000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":["frontiersin.org"],"crossmark-restriction":true},"short-container-title":["Front. Behav. Neurosci."],"abstract":"<jats:sec>\n                    <jats:title>Introduction<\/jats:title>\n                    <jats:p>Applying artificial intelligence to functional magnetic resonance imaging (fMRI) has advanced the modelling of neural activity, but deep architectures such as graph neural networks (GNNs) require large samples, limiting their use in the small cohorts typical of task-based fMRI. Shallow neural networks (SNNs) are robust with few samples but treat brain regions independently, overlooking the brain\u2019s network organization. We propose a structurally constrained message-passing framework that integrates diffusion-based structural connectivity, from a previously published group connectome, with region-level fMRI signals in a shallow architecture.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Methods<\/jats:title>\n                    <jats:p>The task is a binary classification of individual video trials as mentalizing or random, decoded from the fMRI response, in a Theory of Mind paradigm from the Human Connectome Project Young Adult dataset (30 subjects) under subject-wise 3-fold cross-validation. Two shallow classifiers are compared on the same data: the baseline classifies each trial from the regional fMRI signals as independent inputs, whereas the proposed model first applies one message-passing step, replacing each region\u2019s signal with a structurally connected combination of its neighbors\u2019 signals, so that classification operates at the level of anatomically defined networks. Interpretability is assessed using SHAP (SHapley Additive exPlanations).<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Results<\/jats:title>\n                    <jats:p>Across folds, the proposed model matches the baseline (mean retrained accuracy 85.1%; 84.7% on the shared partition), the accuracy lost to pruning being recovered by retraining in every fold. The magnitude of the inputs\u2019 contributions is reproducible across folds, whereas their direction is not. Only about a quarter of the consistently retained inputs also preserve their sign, so a cross-fold sign-consistency criterion is applied to retain reliable explanations. This identifies seven networks contributing robustly to the mentalizing class, located in the temporo-parietal junction, inferior parietal, and medial temporal cortices and overlapping a literature-based mentalizing map (overlap coefficient 0.383). Within these networks, Cohen\u2019s D ranks the relative signal magnitude of the constituent regions.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Discussion<\/jats:title>\n                    <jats:p>Overall, structural constraints improve biological interpretability without compromising classification performance, while cross-validation establishes which explanations are stable, offering a compact and interpretable approach to network-level brain decoding in small cohorts and a basis that may be applied to connectivity disorders.<\/jats:p>\n                  <\/jats:sec>","DOI":"10.3389\/fnbeh.2026.1833995","type":"journal-article","created":{"date-parts":[[2026,9,23]],"date-time":"2026-09-23T06:20:19Z","timestamp":1790144419000},"update-policy":"https:\/\/doi.org\/10.3389\/crossmark-policy","source":"Crossref","is-referenced-by-count":0,"title":["Structurally restricted message-passing within shallow architectures for explainable network-level brain decoding on small cohorts"],"prefix":"10.3389","volume":"20","author":[{"given":"Jos\u00e9 Diogo","family":"Marques dos Santos","sequence":"first","affiliation":[{"name":"Faculty of Engineering, University of Porto","place":["Porto, Portugal"]},{"name":"LIACC \u2013 Artificial Intelligence and Computer Science Laboratory, University of Porto","place":["Porto, Portugal"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Maria Beatriz","family":"Ramos","sequence":"additional","affiliation":[{"name":"Faculty of Medicine, University of Porto","place":["Porto, Portugal"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Lu\u00eds Paulo","family":"Reis","sequence":"additional","affiliation":[{"name":"Faculty of Engineering, University of Porto","place":["Porto, Portugal"]},{"name":"LIACC \u2013 Artificial Intelligence and Computer Science Laboratory, University of Porto","place":["Porto, Portugal"]},{"name":"LASI \u2013 Intelligent Systems Associate Laboratory","place":["Guimar\u00e3es, Portugal"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Jos\u00e9 Paulo","family":"Marques dos Santos","sequence":"additional","affiliation":[{"name":"LIACC \u2013 Artificial Intelligence and Computer Science Laboratory, University of Porto","place":["Porto, Portugal"]},{"name":"Faculty of Medicine, University of Porto","place":["Porto, Portugal"]},{"name":"LASI \u2013 Intelligent Systems Associate Laboratory","place":["Guimar\u00e3es, Portugal"]},{"name":"University of Maia","place":["Maia, Portugal"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Bruno","family":"Direito","sequence":"additional","affiliation":[{"name":"LASI \u2013 Intelligent Systems Associate Laboratory","place":["Guimar\u00e3es, Portugal"]},{"name":"CISUC \u2013 Centre for Informatics and Systems, University of Coimbra","place":["Coimbra, Portugal"]}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"1965","published-online":{"date-parts":[[2026,9,23]]},"reference":[{"key":"ref1","doi-asserted-by":"publisher","first-page":"231","DOI":"10.1016\/S0959-4388(00)00202-6","article-title":"The neurobiology of social cognition","volume":"11","author":"Adolphs","year":"2001","journal-title":"Curr. Opin. Neurobiol."},{"key":"ref2","doi-asserted-by":"publisher","first-page":"S1","DOI":"10.1093\/ons\/opy253","article-title":"A connectomic atlas of the human cerebrum-chapter 1: introduction, methods, and significance","volume":"15","author":"Baker","year":"2018","journal-title":"Oper. Neurosurg."},{"key":"ref3","author":"Baliga","year":"2023"},{"key":"ref4","doi-asserted-by":"publisher","first-page":"353","DOI":"10.1038\/nn.4502","article-title":"Network neuroscience","volume":"20","author":"Bassett","year":"2017","journal-title":"Nat. Neurosci."},{"key":"ref5","doi-asserted-by":"publisher","first-page":"S470","DOI":"10.1093\/ons\/opy272","article-title":"A connectomic atlas of the human cerebrum-chapter 18: the connectional anatomy of human brain networks","volume":"15","author":"Briggs","year":"","journal-title":"Oper. Neurosurg."},{"key":"ref6","doi-asserted-by":"publisher","first-page":"S444","DOI":"10.1093\/ons\/opy268","article-title":"A connectomic atlas of the human cerebrum-chapter 14: Tractographic description of the frontal aslant tract","volume":"15","author":"Briggs","year":"","journal-title":"Oper. Neurosurg."},{"key":"ref7","doi-asserted-by":"publisher","first-page":"S456","DOI":"10.1093\/ons\/opy270","article-title":"A connectomic atlas of the human cerebrum-chapter 16: Tractographic description of the vertical occipital fasciculus","volume":"15","author":"Briggs","year":"","journal-title":"Oper. Neurosurg."},{"key":"ref8","doi-asserted-by":"publisher","first-page":"S462","DOI":"10.1093\/ons\/opy271","article-title":"A connectomic atlas of the human cerebrum-chapter 17: tractographic description of the cingulum","volume":"15","author":"Briggs","year":"","journal-title":"Oper. Neurosurg."},{"key":"ref9","doi-asserted-by":"publisher","first-page":"S450","DOI":"10.1093\/ons\/opy269","article-title":"A connectomic atlas of the human cerebrum-chapter 15: tractographic description of the uncinate fasciculus","volume":"15","author":"Briggs","year":"","journal-title":"Oper. Neurosurg."},{"key":"ref10","doi-asserted-by":"publisher","first-page":"187","DOI":"10.1016\/j.tics.2008.02.010","article-title":"Does the chimpanzee have a theory of mind? 30 years later","volume":"12","author":"Call","year":"2008","journal-title":"Trends Cogn. Sci."},{"key":"ref11","doi-asserted-by":"publisher","first-page":"58","DOI":"10.1007\/s44163-024-00161-0","article-title":"Unleashing the power of advanced technologies for revolutionary medical imaging: pioneering the healthcare frontier with artificial intelligence","volume":"4","author":"Chauhan","year":"2024","journal-title":"Discov. Artif. Intell."},{"key":"ref12","doi-asserted-by":"publisher","first-page":"S407","DOI":"10.1093\/ons\/opy264","article-title":"A connectomic atlas of the human cerebrum-chapter 10: tractographic description of the superior longitudinal fasciculus","volume":"15","author":"Conner","year":"","journal-title":"Oper. Neurosurg."},{"key":"ref13","doi-asserted-by":"publisher","first-page":"S429","DOI":"10.1093\/ons\/opy266","article-title":"A connectomic atlas of the human cerebrum-chapter 12: tractographic description of the middle longitudinal fasciculus","volume":"15","author":"Conner","year":"","journal-title":"Oper. Neurosurg."},{"key":"ref14","doi-asserted-by":"publisher","first-page":"S436","DOI":"10.1093\/ons\/opy267","article-title":"A connectomic atlas of the human cerebrum-chapter 13: tractographic description of the inferior fronto-occipital fasciculus","volume":"15","author":"Conner","year":"","journal-title":"Oper. Neurosurg."},{"key":"ref15","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1080\/17470910600683549","article-title":"Social neuroscience: a new journal","volume":"1","author":"Decety","year":"2006","journal-title":"Soc. Neurosci."},{"key":"ref16","author":"Doran","year":"2018"},{"key":"ref17","doi-asserted-by":"publisher","first-page":"525","DOI":"10.3389\/fnins.2018.00525","article-title":"Classification and prediction of brain disorders using functional connectivity: promising but challenging","volume":"12","author":"Du","year":"2018","journal-title":"Front. Neurosci."},{"key":"ref18","doi-asserted-by":"publisher","first-page":"e86327","DOI":"10.7554\/eLife.86327","article-title":"Gaze patterns and brain activations in humans and marmosets in the Frith-Happ\u00e9 theory-of-mind animation task","volume":"12","author":"Dureux","year":"2023","journal-title":"eLife"},{"key":"ref19","doi-asserted-by":"publisher","first-page":"118543","DOI":"10.1016\/j.neuroimage.2021.118543","article-title":"The human connectome project: a retrospective","volume":"244","author":"Elam","year":"2021","journal-title":"NeuroImage"},{"key":"ref20","doi-asserted-by":"publisher","first-page":"906290","DOI":"10.3389\/fnins.2022.906290","article-title":"Explainable AI: a review of applications to neuroimaging data [systematic review]","volume":"16","author":"Farahani","year":"2022","journal-title":"Front. Neurosci."},{"key":"ref22","doi-asserted-by":"publisher","first-page":"171","DOI":"10.1038\/nature18933","article-title":"A multi-modal parcellation of human cerebral cortex","volume":"536","author":"Glasser","year":"2016","journal-title":"Nature"},{"key":"ref23","volume-title":"Neural Networks and Learning Machines","author":"Haykin","year":"2009"},{"key":"ref24","doi-asserted-by":"publisher","first-page":"243","DOI":"10.2307\/1416950","article-title":"An experimental study of apparent behavior","volume":"57","author":"Heider","year":"1944","journal-title":"Am. J. Psychol."},{"key":"ref25","doi-asserted-by":"publisher","first-page":"176","DOI":"10.1016\/j.cortex.2016.04.017","article-title":"Social representations and contextual adjustments as two distinct components of the theory of mind brain network: evidence from the REMICS task","volume":"81","author":"Lavoie","year":"2016","journal-title":"Cortex"},{"key":"ref26","author":"Limas","year":"2014"},{"key":"ref27","author":"Lundberg","year":"2017"},{"key":"ref28","doi-asserted-by":"crossref","first-page":"260","DOI":"10.1007\/978-3-031-07802-6_22","volume-title":"Bioinformatics and Biomedical Engineering. Lecture Notes in Computer Science","author":"Marques dos Santos","year":"2022"},{"key":"ref29","first-page":"417","volume-title":"Machine Learning, Optimization, and Data Science. Lecture Notes in Computer Science","author":"Marques dos Santos","year":"2023"},{"key":"ref30","doi-asserted-by":"publisher","first-page":"17","DOI":"10.3390\/make7010017","article-title":"Decoding mental states in social cognition: insights from explainable artificial intelligence on HCP fMRI data","volume":"7","author":"Marques dos Santos","year":"2025","journal-title":"Mach. Learn. Knowl. Extr."},{"key":"ref31","doi-asserted-by":"publisher","first-page":"1305164","DOI":"10.3389\/fnhum.2024.1305164","article-title":"XAI (explainable artificial intelligence) in neuromarketing \/ consumer neuroscience: an fMRI study on brand perception","volume":"18","author":"Marques dos Santos","year":"2024","journal-title":"Front. Hum. Neurosci."},{"key":"ref32","doi-asserted-by":"publisher","first-page":"17","DOI":"10.3390\/brainsci15010017","article-title":"Graph neural networks in brain connectivity studies: methods, challenges, and future directions","volume":"15","author":"Mohammadi","year":"2025","journal-title":"Brain Sci."},{"key":"ref33","doi-asserted-by":"publisher","first-page":"60","DOI":"10.1002\/wcs.29","article-title":"Social neuroscience","volume":"1","author":"Norman","year":"2010","journal-title":"Wiley Interdiscip. Rev. Cogn. Sci."},{"key":"ref34","doi-asserted-by":"publisher","first-page":"137267","DOI":"10.1016\/j.neulet.2023.137267","article-title":"Heterogeneity of social cognition between visual perspective-taking and theory of mind in the temporo-parietal junction","volume":"807","author":"Ogawa","year":"2023","journal-title":"Neurosci. Lett."},{"key":"ref35","doi-asserted-by":"publisher","first-page":"150","DOI":"10.1016\/j.neulet.2009.07.064","article-title":"Individual differences in the theory of mind and superior temporal sulcus","volume":"463","author":"Otsuka","year":"2009","journal-title":"Neurosci. Lett."},{"key":"ref36","doi-asserted-by":"publisher","first-page":"S199","DOI":"10.1016\/j.neuroimage.2008.11.007","article-title":"Machine learning classifiers and fMRI: a tutorial overview","volume":"45","author":"Pereira","year":"2009","journal-title":"NeuroImage"},{"key":"ref37","doi-asserted-by":"publisher","first-page":"515","DOI":"10.1017\/S0140525X00076512","article-title":"Does the chimpanzee have a theory of mind?","volume":"1","author":"Premack","year":"1978","journal-title":"Behav. Brain Sci."},{"key":"ref38","doi-asserted-by":"publisher","first-page":"692","DOI":"10.3390\/brainsci13040692","article-title":"Increased functional connectivity involving the parahippocampal gyrus in patients with schizophrenia during theory of mind processing: a psychophysiological interaction study","volume":"13","author":"Qin","year":"2023","journal-title":"Brain Sci."},{"key":"ref39","doi-asserted-by":"publisher","first-page":"178","DOI":"10.1016\/j.biopsych.2022.05.031","article-title":"Shared and specific patterns of structural brain connectivity across affective and psychotic disorders","volume":"93","author":"Repple","year":"2023","journal-title":"Biol. Psychiatry"},{"key":"ref40","doi-asserted-by":"publisher","first-page":"42200","DOI":"10.1109\/ACCESS.2020.2976199","article-title":"Explainable machine learning for scientific insights and discoveries","volume":"8","author":"Roscher","year":"2020","journal-title":"IEEE Access"},{"key":"ref41","doi-asserted-by":"publisher","first-page":"S423","DOI":"10.1093\/ons\/opy265","article-title":"A connectomic atlas of the human cerebrum-chapter 11: tractographic description of the inferior longitudinal fasciculus","volume":"15","author":"Sali","year":"2018","journal-title":"Oper. Neurosurg. (Hagerstown, Md.)"},{"key":"ref42","first-page":"83","volume-title":"Social Neuroscience: People Thinking about Thinking People","author":"Saxe","year":"2006"},{"key":"ref43","doi-asserted-by":"publisher","first-page":"9","DOI":"10.1016\/j.neubiorev.2014.01.009","article-title":"Fractionating theory of mind: a meta-analysis of functional brain imaging studies","volume":"42","author":"Schurz","year":"2014","journal-title":"Neurosci. Biobehav. Rev."},{"key":"ref44","doi-asserted-by":"publisher","first-page":"617","DOI":"10.1093\/brain\/awn279","article-title":"Two systems for empathy: a double dissociation between emotional and cognitive empathy in inferior frontal gyrus versus ventromedial prefrontal lesions","volume":"132","author":"Shamay-Tsoory","year":"2009","journal-title":"Brain"},{"key":"ref45","author":"Shrikumar","year":"2017"},{"key":"ref46","doi-asserted-by":"publisher","first-page":"nsae014","DOI":"10.1093\/scan\/nsae014","article-title":"Deep social neuroscience: the promise and peril of using artificial neural networks to study the social brain","volume":"19","author":"Sievers","year":"2024","journal-title":"Soc. Cogn. Affect. Neurosci."},{"key":"ref47","doi-asserted-by":"crossref","first-page":"513","DOI":"10.1016\/B978-0-12-416008-8.00027-9","volume-title":"Neuroeconomics: Decision Making and the Brain","author":"Singer","year":"2014"},{"key":"ref49","first-page":"cer-10-16","article-title":"The human connectome project: progress and prospects","volume":"2016","author":"Van Essen","year":"2016","journal-title":"Cerebrum"},{"key":"ref50","doi-asserted-by":"publisher","first-page":"62","DOI":"10.1016\/j.neuroimage.2013.05.041","article-title":"The WU-Minn human connectome project: an overview","volume":"80","author":"Van Essen","year":"2013","journal-title":"NeuroImage"},{"key":"ref51","doi-asserted-by":"publisher","first-page":"1997","DOI":"10.3390\/make6030098","article-title":"Tertiary review on explainable artificial intelligence: where do we stand?","volume":"6","author":"Van Mourik","year":"2024","journal-title":"Mach. Learn. Knowl. Extr."},{"key":"ref52","doi-asserted-by":"publisher","first-page":"615","DOI":"10.3390\/make3030032","article-title":"Classification of explainable artificial intelligence methods through their output formats","volume":"3","author":"Vilone","year":"2021","journal-title":"Mach. Learn. Knowl. Extr."},{"key":"ref53","doi-asserted-by":"publisher","first-page":"23","DOI":"10.3389\/fninf.2018.00023","article-title":"Deep learning methods to process fMRI data and their application in the diagnosis of cognitive impairment: a brief overview and our opinion","volume":"12","author":"Wen","year":"2018","journal-title":"Front. Neuroinform."},{"key":"ref9001","doi-asserted-by":"publisher","DOI":"10.48550\/arxiv.1902.07153","article-title":"Simplifying graph convolutional networks","author":"Wu","year":"2019","journal-title":"arXiv (Cornell University)"},{"key":"ref54","doi-asserted-by":"publisher","first-page":"2658","DOI":"10.1016\/j.neuropsychologia.2010.05.012","article-title":"What gets the attention of the temporo-parietal junction? An fMRI investigation of attention and theory of mind","volume":"48","author":"Young","year":"2010","journal-title":"Neuropsychologia"}],"container-title":["Frontiers in Behavioral Neuroscience"],"original-title":[],"link":[{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/fnbeh.2026.1833995\/full","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,9,23]],"date-time":"2026-09-23T06:20:21Z","timestamp":1790144421000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/fnbeh.2026.1833995\/full"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,9,23]]},"references-count":53,"alternative-id":["10.3389\/fnbeh.2026.1833995"],"URL":"https:\/\/doi.org\/10.3389\/fnbeh.2026.1833995","relation":{},"ISSN":["1662-5153"],"issn-type":[{"value":"1662-5153","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,9,23]]},"article-number":"1833995"}}