{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,5,8]],"date-time":"2026-05-08T06:17:12Z","timestamp":1778221032904,"version":"3.51.4"},"reference-count":51,"publisher":"Frontiers Media SA","license":[{"start":{"date-parts":[[2026,5,8]],"date-time":"2026-05-08T00:00:00Z","timestamp":1778198400000},"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. Neurosci."],"abstract":"<jats:sec>\n                    <jats:title>Introduction<\/jats:title>\n                    <jats:p>Accurate and reliable detection of speech state transitions is a prerequisite for practical speech brain\u2013computer interfaces (BCIs). While cortical language areas have been extensively studied, it remains unclear whether speech onset information is exclusively localized to these regions or distributed across a broader cortico-subcortical network. Here, we investigated the feasibility of decoding speech state transitions using sparse stereo-electroencephalography (SEEG) recordings that sample both cortical and subcortical structures.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Methods<\/jats:title>\n                    <jats:p>Four Mandarin-speaking epilepsy patients undergoing clinical SEEG monitoring performed a sentence-reading task. Neural signals were segmented and labeled as rest or speech based on acoustic onset. A convolutional neural network was trained to classify speech states using broadband or high-gamma features derived from different anatomical channel subsets. We further evaluated continuous decoding performance, model robustness to channel dropout, and the specific contributions of different brain regions.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Results<\/jats:title>\n                    <jats:p>Speech state decoding accuracy exceeded chance level (50%) in all participants, with peak single-participant accuracies surpassing 90%. Models integrating both cortical and subcortical signals generally outperformed those restricted to a single anatomical domain. Notably, broadband signals yielded higher classification accuracy than high-gamma features. In continuous decoding simulations, performance remained above chance, although reduced relative to discretized evaluation. Crucially, decoding accuracy was robust to random channel reduction (up to 50%) and remained above 70% even after excluding classical speech-related cortical regions. Contribution analyses indicated participant-specific patterns of model sensitivity, with relatively higher contributions observed in frontal regions and the thalamus in multiple participants.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Discussion<\/jats:title>\n                    <jats:p>These findings support the hypothesis that speech state information is represented in a distributed cortico-subcortical network rather than being confined to canonical language areas. The robustness of decoding performance despite channel reduction and regional exclusion suggests that sparsely sampled SEEG data can effectively drive speech detection modules. This study demonstrates the feasibility of utilizing deep brain recordings for speech BCIs, offering a pathway toward more stable and generalized implantable systems. Moreover, such autonomous speech state detection may also serve as an ethical safeguard, ensuring that neural language decoding is activated only during intended communicative acts.<\/jats:p>\n                  <\/jats:sec>","DOI":"10.3389\/fnins.2026.1816455","type":"journal-article","created":{"date-parts":[[2026,5,8]],"date-time":"2026-05-08T05:56:28Z","timestamp":1778219788000},"update-policy":"https:\/\/doi.org\/10.3389\/crossmark-policy","source":"Crossref","is-referenced-by-count":0,"title":["Distributed cortico-subcortical networks enable robust speech state detection from sparse intracranial recordings"],"prefix":"10.3389","volume":"20","author":[{"given":"Chen","family":"Feng","sequence":"first","affiliation":[{"name":"College of Biomedical Engineering and Instrument Science, Zhejiang University","place":["Hangzhou, China"]},{"name":"Department of Neurosurgery, The Second Affiliated Hospital Zhejiang University School of Medicine","place":["Hangzhou, China"]}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"En","family":"Zhang","sequence":"additional","affiliation":[{"name":"State Key Laboratory of Cognitive Neuroscience and Learning and IDG\/McGovern Institute for Brain Research, Beijing Normal University","place":["Beijing, China"]}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Yifei","family":"Jia","sequence":"additional","affiliation":[{"name":"School of Engineering, Westlake University","place":["Hangzhou, China"]}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Zhoule","family":"Zhu","sequence":"additional","affiliation":[{"name":"Department of Neurosurgery, The Second Affiliated Hospital Zhejiang University School of Medicine","place":["Hangzhou, China"]}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Junming","family":"Zhu","sequence":"additional","affiliation":[{"name":"Department of Neurosurgery, The Second Affiliated Hospital Zhejiang University School of Medicine","place":["Hangzhou, China"]}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Di","family":"Wu","sequence":"additional","affiliation":[{"name":"School of Intelligence Science and Engineering, College of Artificial Intelligence, Harbin Institute of Technology","place":["Shenzhen, China"]}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Kedi","family":"Xu","sequence":"additional","affiliation":[{"name":"College of Biomedical Engineering and Instrument Science, Zhejiang University","place":["Hangzhou, China"]}],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"1965","published-online":{"date-parts":[[2026,5,8]]},"reference":[{"key":"ref1","doi-asserted-by":"publisher","first-page":"1055","DOI":"10.1038\/s42003-021-02578-0","article-title":"Real-time synthesis of imagined speech processes from minimally invasive recordings of neural activity","volume":"4","author":"Angrick","year":"2021","journal-title":"Commun. Biol."},{"key":"ref2","doi-asserted-by":"publisher","first-page":"493","DOI":"10.1038\/s41586-019-1119-1","article-title":"Speech synthesis from neural decoding of spoken sentences","volume":"568","author":"Anumanchipalli","year":"2019","journal-title":"Nature"},{"key":"ref3","first-page":"85","article-title":"Functional stereotaxic exploration (Seeg) of epilepsy","volume":"28","author":"Bancaud","year":"1970","journal-title":"Electroencephalogr. Clin. Neurophysiol."},{"key":"ref4","doi-asserted-by":"publisher","first-page":"1131","DOI":"10.1111\/epi.13791","article-title":"Defining epileptogenic networks: contribution of Seeg and signal analysis","volume":"58","author":"Bartolomei","year":"2017","journal-title":"Epilepsia"},{"key":"ref5","doi-asserted-by":"publisher","first-page":"53","DOI":"10.1016\/j.neucli.2017.11.010","article-title":"Interpretation of Seeg recordings","volume":"48","author":"Bartolomei","year":"2018","journal-title":"Neurophysiol. Clin."},{"key":"ref6","doi-asserted-by":"publisher","first-page":"271","DOI":"10.1038\/s42003-025-07464-7","article-title":"Learning to operate an imagined speech brain-computer Interface involves the spatial and frequency tuning of neural activity","volume":"8","author":"Bhadra","year":"2025","journal-title":"Commun. Biol."},{"key":"ref7","article-title":"Praat: doing phonetics by computer [Computer program]","author":"Boersma","year":"2011"},{"key":"ref8","first-page":"330","article-title":"Remarques sur le si\u00e8ge de la facult\u00e9 du langage articul\u00e9, suivies d\u2019une observation d\u2019aph\u00e9mie (perte de la parole)","volume":"6","author":"Broca","year":"1861","journal-title":"Bulletin et Memoires de la Societe anatomique de Paris"},{"key":"ref9","doi-asserted-by":"publisher","first-page":"118962","DOI":"10.1016\/j.neuroimage.2022.118962","article-title":"Differentiation of speech-induced artifacts from physiological high gamma activity in intracranial recordings","volume":"250","author":"Bush","year":"2022","journal-title":"NeuroImage"},{"key":"ref10","doi-asserted-by":"publisher","first-page":"203","DOI":"10.1146\/annurev-neuro-062111-150444","article-title":"Mechanisms of gamma oscillations","volume":"35","author":"Buzsaki","year":"2012","journal-title":"Annu. Rev. Neurosci."},{"key":"ref11","doi-asserted-by":"publisher","first-page":"609","DOI":"10.1056\/NEJMoa2314132","article-title":"An accurate and rapidly calibrating speech neuroprosthesis","volume":"391","author":"Card","year":"2024","journal-title":"N. Engl. J. Med."},{"key":"ref12","doi-asserted-by":"publisher","first-page":"2271","DOI":"10.1093\/brain\/121.12.2271","article-title":"Functional mapping of human sensorimotor cortex with electrocorticographic spectral analysis. I. Alpha and beta event-related desynchronization","volume":"121","author":"Crone","year":"1998","journal-title":"Brain"},{"key":"ref13","doi-asserted-by":"publisher","first-page":"73","DOI":"10.1016\/j.bandl.2012.06.011","article-title":"Thalamic mechanisms in language: a reconsideration based on recent findings and concepts","volume":"126","author":"Crosson","year":"2013","journal-title":"Brain Lang."},{"key":"ref14","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1016\/j.bandl.2013.05.011","article-title":"A re-examination of neural basis of language processing: proposal of a dynamic hodotopical model from data provided by brain stimulation mapping during picture naming","volume":"131","author":"Duffau","year":"2014","journal-title":"Brain Lang."},{"key":"ref15","doi-asserted-by":"publisher","first-page":"0257","DOI":"10.34133\/cbsystems.0257","article-title":"Acoustic inspired brain-to-sentence decoder for logosyllabic language","volume":"6","author":"Feng","year":"2025","journal-title":"Cyborg Bionic Syst."},{"key":"ref16","doi-asserted-by":"publisher","first-page":"774","DOI":"10.1016\/j.neuroimage.2012.01.021","article-title":"FreeSurfer","volume":"62","author":"Fischl","year":"2012","journal-title":"Neuroimage"},{"key":"ref17","doi-asserted-by":"publisher","first-page":"220","DOI":"10.1016\/j.neuron.2015.09.034","article-title":"Rhythms for cognition: communication through coherence","volume":"88","author":"Fries","year":"2015","journal-title":"Neuron"},{"key":"ref18","doi-asserted-by":"publisher","first-page":"940","DOI":"10.1126\/science.170.3961.940","article-title":"The Organization of Language and the brain: language disorders after brain damage help in elucidating the neural basis of verbal behavior","volume":"170","author":"Geschwind","year":"1970","journal-title":"Science"},{"key":"ref19","doi-asserted-by":"publisher","first-page":"57","DOI":"10.1016\/B978-0-12-823384-9.00003-7","article-title":"The dual stream model of speech and language processing","volume":"185","author":"Hickok","year":"2022","journal-title":"Handb. Clin. Neurol."},{"key":"ref20","doi-asserted-by":"publisher","first-page":"393","DOI":"10.1038\/nrn2113","article-title":"The cortical organization of speech processing","volume":"8","author":"Hickok","year":"2007","journal-title":"Nat. Rev. Neurosci."},{"key":"ref21","doi-asserted-by":"publisher","first-page":"174","DOI":"10.1016\/j.conb.2015.04.006","article-title":"The cortical computations underlying feedback control in vocal production","volume":"33","author":"Houde","year":"2015","journal-title":"Curr. Opin. Neurobiol."},{"key":"ref22","doi-asserted-by":"publisher","first-page":"982","DOI":"10.1016\/j.cortex.2009.02.010","article-title":"Non-motor basal ganglia functions: a review and proposal for a model of sensory predictability in auditory language perception","volume":"45","author":"Kotz","year":"2009","journal-title":"Cortex"},{"key":"ref23","doi-asserted-by":"publisher","first-page":"4658","DOI":"10.1016\/j.cell.2025.06.015","article-title":"Inner speech in motor cortex and implications for speech neuroprostheses","volume":"188","author":"Kunz","year":"2025","journal-title":"Cell"},{"key":"ref24","doi-asserted-by":"publisher","first-page":"593","DOI":"10.1038\/s41586-023-06839-2","article-title":"Large-scale single-neuron speech sound encoding across the depth of human cortex","volume":"626","author":"Leonard","year":"2024","journal-title":"Nature"},{"key":"ref25","doi-asserted-by":"publisher","first-page":"1442","DOI":"10.1038\/s42003-025-08847-6","article-title":"Evaluating the temporal order of motor and auditory systems in speech production using intracranial Eeg","volume":"8","author":"Li","year":"2025","journal-title":"Commun. Biol."},{"key":"ref26","doi-asserted-by":"publisher","first-page":"902","DOI":"10.1038\/s41593-025-01905-6","article-title":"A streaming brain-to-voice neuroprosthesis to restore naturalistic communication","volume":"28","author":"Littlejohn","year":"2025","journal-title":"Nat. Neurosci."},{"key":"ref27","doi-asserted-by":"publisher","first-page":"98","DOI":"10.1038\/s41586-021-04191-x","article-title":"Transforming representations of movement from body- to world-centric space","volume":"601","author":"Lu","year":"2022","journal-title":"Nature"},{"key":"ref28","volume-title":"Rsi-Grad-Cam: Visual Explanations from Deep Networks via Riemann-Stieltjes Integrated Gradient-Based Localization. In: Bebis, G., et al. Advances in Visual Computing. ISVC 2022. Lecture Notes in Computer Science, vol 13598. Cham: Springer","author":"Lucas","year":""},{"key":"ref29","doi-asserted-by":"publisher","first-page":"1293","DOI":"10.1098\/rstb.2001.0915","article-title":"A probabilistic atlas and reference system for the human brain: international consortium for brain mapping (Icbm)","volume":"356","author":"Mazziotta","year":"2001","journal-title":"Philos. Trans. R. Soc. Lond. Ser. B Biol. Sci."},{"key":"ref30","doi-asserted-by":"publisher","first-page":"1037","DOI":"10.1038\/s41586-023-06443-4","article-title":"A high-performance neuroprosthesis for speech decoding and avatar control","volume":"620","author":"Metzger","year":"2023","journal-title":"Nature"},{"key":"ref31","doi-asserted-by":"publisher","first-page":"217","DOI":"10.1056\/NEJMoa2027540","article-title":"Neuroprosthesis for decoding speech in a paralyzed person with anarthria","volume":"385","author":"Moses","year":"2021","journal-title":"N. Engl. J. Med."},{"key":"ref32","doi-asserted-by":"publisher","first-page":"91","DOI":"10.1016\/0165-0173(94)00007-C","article-title":"Functional anatomy of the basal ganglia. I. The cortico-basal ganglia-thalamo-cortical loop","volume":"20","author":"Parent","year":"1995","journal-title":"Brain Res. Brain Res. Rev."},{"key":"ref33","doi-asserted-by":"publisher","first-page":"816","DOI":"10.1016\/j.neuroimage.2012.04.062","article-title":"A review and synthesis of the first 20 years of pet and fmri studies of heard speech, spoken language and reading","volume":"62","author":"Price","year":"2012","journal-title":"NeuroImage"},{"key":"ref34","doi-asserted-by":"publisher","first-page":"056028","DOI":"10.1088\/1741-2552\/abb25e","article-title":"Observation and assessment of acoustic contamination of electrophysiological brain signals during speech production and sound perception","volume":"17","author":"Roussel","year":"2020","journal-title":"J. Neural Eng."},{"key":"ref35","doi-asserted-by":"publisher","first-page":"575","DOI":"10.1016\/j.conb.2012.03.005","article-title":"Thalamocortical interactions","volume":"22","author":"Sherman","year":"2012","journal-title":"Curr. Opin. Neurobiol."},{"key":"ref36","doi-asserted-by":"publisher","first-page":"121","DOI":"10.1038\/nrn3137","article-title":"Spectral fingerprints of large-scale neuronal interactions","volume":"13","author":"Siegel","year":"2012","journal-title":"Nat. Rev. Neurosci."},{"key":"ref37","doi-asserted-by":"publisher","first-page":"473","DOI":"10.1038\/s41583-024-00819-9","article-title":"The speech neuroprosthesis","volume":"25","author":"Silva","year":"","journal-title":"Nat. Rev. Neurosci."},{"key":"ref38","doi-asserted-by":"publisher","first-page":"977","DOI":"10.1038\/s41551-024-01207-5","article-title":"A bilingual speech neuroprosthesis driven by cortical articulatory representations shared between languages","volume":"8","author":"Silva","year":"","journal-title":"Nat. Biomed. Eng."},{"key":"ref39","doi-asserted-by":"publisher","first-page":"282","DOI":"10.1007\/s12311-020-01207-6","article-title":"Contribution of the cerebellum and the basal ganglia to language production: speech, word fluency, and sentence construction-evidence from pathology","volume":"20","author":"Silveri","year":"2021","journal-title":"Cerebellum"},{"key":"ref40","doi-asserted-by":"publisher","first-page":"119913","DOI":"10.1016\/j.neuroimage.2023.119913","article-title":"The nested hierarchy of overt, mouthed, and imagined speech activity evident in intracranial recordings","volume":"269","author":"Soroush","year":"2023","journal-title":"NeuroImage"},{"key":"ref41","doi-asserted-by":"publisher","first-page":"100017","DOI":"10.59717\/j.xinn-inform.2025.100017","article-title":"Psyrtdevice: open-source toolkit for sub-millisecond evaluation of reaction time device accuracy","volume":"1","author":"Tang","year":"2025","journal-title":"Innov. Informat."},{"key":"ref42","doi-asserted-by":"publisher","first-page":"952","DOI":"10.1080\/01690960903498424","article-title":"The diva model: a neural theory of speech acquisition and production","volume":"26","author":"Tourville","year":"2011","journal-title":"Lang. Cogn. Process."},{"key":"ref43","doi-asserted-by":"publisher","first-page":"432","DOI":"10.1038\/s42003-025-07862-x","article-title":"Whole-brain dynamics of articulatory, acoustic and semantic speech representations","volume":"8","author":"Verwoert","year":"2025","journal-title":"Commun. Biol."},{"key":"ref44","doi-asserted-by":"publisher","first-page":"434","DOI":"10.1038\/s41597-022-01542-9","article-title":"Dataset of speech production in intracranial electroencephalography","volume":"9","author":"Verwoert","year":"2022","journal-title":"Sci. Data"},{"key":"ref45","doi-asserted-by":"publisher","first-page":"145","DOI":"10.1038\/s41586-025-09127-3","article-title":"An instantaneous voice-synthesis neuroprosthesis","volume":"644","author":"Wairagkar","year":"2025","journal-title":"Nature"},{"key":"ref46","doi-asserted-by":"publisher","first-page":"111432","DOI":"10.1016\/j.brainresbull.2025.111432","article-title":"The unique role of the associative thalamus in cognitive processing","volume":"229","author":"Wang","year":"2025","journal-title":"Brain Res. Bull."},{"key":"ref47","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.2302.14367","article-title":"BrainBERT: self-supervised representation learning for intracranial recordings","author":"Wang","year":"2023","journal-title":"ArXiv"},{"key":"ref48","doi-asserted-by":"publisher","first-page":"1031","DOI":"10.1038\/s41586-023-06377-x","article-title":"A high-performance speech neuroprosthesis","volume":"620","author":"Willett","year":"2023","journal-title":"Nature"},{"key":"ref49","doi-asserted-by":"publisher","first-page":"222","DOI":"10.1016\/j.neubiorev.2020.11.013","article-title":"A thalamic bridge from sensory perception to cognition","volume":"120","author":"Wolff","year":"2021","journal-title":"Neurosci. Biobehav. Rev."},{"key":"ref50","doi-asserted-by":"publisher","first-page":"036055","DOI":"10.1088\/1741-2552\/ad593a","article-title":"Speech decoding from stereo-electroencephalography (seeg) signals using advanced deep learning methods","volume":"21","author":"Wu","year":"2024","journal-title":"J. Neural Eng."},{"key":"ref51","doi-asserted-by":"publisher","first-page":"159","DOI":"10.1038\/551159a","article-title":"Four ethical priorities for neurotechnologies and Ai","volume":"551","author":"Yuste","year":"2017","journal-title":"Nature"}],"container-title":["Frontiers in Neuroscience"],"original-title":[],"link":[{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/fnins.2026.1816455\/full","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,5,8]],"date-time":"2026-05-08T05:56:31Z","timestamp":1778219791000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/fnins.2026.1816455\/full"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,5,8]]},"references-count":51,"alternative-id":["10.3389\/fnins.2026.1816455"],"URL":"https:\/\/doi.org\/10.3389\/fnins.2026.1816455","relation":{},"ISSN":["1662-453X"],"issn-type":[{"value":"1662-453X","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,5,8]]},"article-number":"1816455"}}