{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,6,24]],"date-time":"2026-06-24T12:55:27Z","timestamp":1782305727733,"version":"3.54.5"},"reference-count":53,"publisher":"IOP Publishing","issue":"2","license":[{"start":{"date-parts":[[2026,4,7]],"date-time":"2026-04-07T00:00:00Z","timestamp":1775520000000},"content-version":"vor","delay-in-days":6,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"},{"start":{"date-parts":[[2026,4,7]],"date-time":"2026-04-07T00:00:00Z","timestamp":1775520000000},"content-version":"tdm","delay-in-days":6,"URL":"https:\/\/iopscience.iop.org\/info\/page\/text-and-data-mining"},{"start":{"date-parts":[[2026,4,7]],"date-time":"2026-04-07T00:00:00Z","timestamp":1775520000000},"content-version":"vor","delay-in-days":6,"URL":"https:\/\/publishingsupport.iopscience.iop.org\/questions\/alternative-author-rights-policies\/"},{"start":{"date-parts":[[2026,4,7]],"date-time":"2026-04-07T00:00:00Z","timestamp":1775520000000},"content-version":"vor","delay-in-days":6,"URL":"https:\/\/publishingsupport.iopscience.iop.org\/questions\/alternative-author-rights-policies\/"}],"funder":[{"DOI":"10.13039\/501100020963","name":"Moonshot Research and Development Program","doi-asserted-by":"crossref","award":["JPMJMS2012"],"award-info":[{"award-number":["JPMJMS2012"]}],"id":[{"id":"10.13039\/501100020963","id-type":"DOI","asserted-by":"crossref"}]}],"content-domain":{"domain":["iopscience.iop.org"],"crossmark-restriction":false},"short-container-title":["J. Neural Eng."],"published-print":{"date-parts":[[2026,4,1]]},"abstract":"<jats:title>Abstract<\/jats:title>\n                  <jats:p>\n                    <jats:italic>Objective<\/jats:italic>\n                    . Silent speech decoding (SSD) offers a potential communication alternative for individuals with impaired vocalization. However, conventional multi-electrode electroencephalography (EEG) or facial electromyography (EMG) systems require cumbersome preparation and are unsuitable for daily use. This study evaluates the practicality of SSD using a wearable around-ear EEG device, focusing on data scaling, cross-subject transfer, vocabulary extensibility, and online decoding performance.\n                    <jats:italic>Approach<\/jats:italic>\n                    . We collected 72\u2009h of around-ear EEG from 24 healthy participants and one individual with incomplete locked-in syndrome (LIS) during silent, vocalized, and attempted speech, and integrated these around-ear EEG recordings with prior EMG + high-density EEG datasets, yielding 282.4 total h of training data. Using a 64-word classification task as the evaluation metric, we assessed: (1) whether larger datasets improve around-ear EEG\u2013based SSD, (2) whether healthy-participant data supplement limited LIS-participant data despite articulatory differences, (3) transferability to unseen vocabulary, and (4) online user-interface performance.\n                    <jats:italic>Main results<\/jats:italic>\n                    . Large-scale EEG\/EMG data improved SSD accuracy in both healthy participants and the LIS participant. Training on the heterogeneous dataset achieved 56.6% accuracy for healthy users and 47.3% for the LIS participant. Fine-tuning this decoder for new vocabulary increased the accuracy by 22 percentage points relative to training from scratch. Regression analysis showed that, for decoding in the LIS participant, data from the LIS participant contributed approximately four times the weight of healthy-participant data, quantifying data strategies for SSD. Online experiments achieved top-1\/top-5 accuracies of 47.2%\/76.0% for healthy users and 26.5%\/49.1% for the LIS participant.\n                    <jats:italic>Significance<\/jats:italic>\n                    . The results indicate that lightweight, commercially feasible around-ear EEG can enable practical SSD when combined with large-scale healthy-participant data, supporting online operation. Moreover, models trained on a 64-word vocabulary facilitate decoding of a new vocabulary, providing a path toward SSD systems requiring minimal LIS-participant data. This study advances non-invasive SSD systems suitable for everyday communication.\n                  <\/jats:p>","DOI":"10.1088\/1741-2552\/ae54d0","type":"journal-article","created":{"date-parts":[[2026,3,19]],"date-time":"2026-03-19T22:52:23Z","timestamp":1773960743000},"page":"026027","update-policy":"https:\/\/doi.org\/10.1088\/crossmark-policy","source":"Crossref","is-referenced-by-count":2,"title":["Large-scale training data enhances silent speech decoding with around-ear EEG"],"prefix":"10.1088","volume":"23","author":[{"ORCID":"https:\/\/orcid.org\/0009-0009-2888-936X","authenticated-orcid":true,"given":"Masakazu","family":"Inoue","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0009-0001-8285-9091","authenticated-orcid":true,"given":"Eri","family":"Hatakeyama","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Yuya","family":"Kita","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-9941-6510","authenticated-orcid":true,"given":"Shuntaro","family":"Sasai","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"266","published-online":{"date-parts":[[2026,4,7]]},"reference":[{"key":"jneae54d0bib1","doi-asserted-by":"publisher","first-page":"387","DOI":"10.1007\/978-3-030-41299-9_30","type":"conference-proceedings","article-title":"EEG representations of spatial and temporal features in imagined speech and overt speech","author":"Lee","year":"2019"},{"key":"jneae54d0bib2","doi-asserted-by":"publisher","first-page":"6030","DOI":"10.1609\/aaai.v37i5.25745","type":"conference-proceedings","article-title":"Towards voice reconstruction from EEG during imagined speech","volume":"vol 37","author":"Lee","year":"2023"},{"key":"jneae54d0bib3","doi-asserted-by":"publisher","first-page":"2386","DOI":"10.1109\/TASLP.2017.2740000","type":"journal-article","article-title":"Silent speech recognition as an alternative communication device for persons with laryngectomy","volume":"25","author":"Meltzner","year":"2017","journal-title":"IEEE\/ACM Trans. Audio Speech Lang. Process."},{"key":"jneae54d0bib4","doi-asserted-by":"publisher","first-page":"341","DOI":"10.1016\/j.specom.2009.12.002","type":"journal-article","article-title":"Modeling coarticulation in EMG-based continuous speech recognition","volume":"52","author":"Schultz","year":"2010","journal-title":"Speech Commun."},{"key":"jneae54d0bib5","doi-asserted-by":"publisher","first-page":"43","DOI":"10.1145\/3172944.3172977","type":"conference-proceedings","article-title":"Alterego: a personalized wearable silent speech interface","author":"Kapur","year":"2018"},{"key":"jneae54d0bib6","article-title":"A cross-modal approach to silent speech with LLM-enhanced recognition","author":"Benster","year":"2024","type":"preprint"},{"key":"jneae54d0bib7","doi-asserted-by":"publisher","first-page":"1824","DOI":"10.1109\/BigData62323.2024.10825980","type":"conference-proceedings","article-title":"Eeg2text: open vocabulary EEG-to-text translation with multi-view transformer","author":"Liu","year":"2024"},{"key":"jneae54d0bib8","doi-asserted-by":"publisher","DOI":"10.1088\/1741-2552\/ad8d0a","type":"journal-article","article-title":"Continuous and discrete decoding of overt speech with scalp electroencephalography (EEG)","volume":"22","author":"Craik","year":"2025","journal-title":"J. Neural Eng."},{"key":"jneae54d0bib9","article-title":"Voicing silent speech","author":"Gaddy","year":"2022","type":"other"},{"key":"jneae54d0bib10","doi-asserted-by":"publisher","first-page":"20","DOI":"10.1038\/s41528-021-00119-7","type":"journal-article","article-title":"All-weather, natural silent speech recognition via machine-learning-assisted tattoo-like electronics","volume":"5","author":"Wang","year":"2021","journal-title":"npj Flex. Electron."},{"key":"jneae54d0bib11","doi-asserted-by":"publisher","DOI":"10.1016\/j.snr.2025.100407","type":"journal-article","article-title":"A silent speech interface with machine learning recognition model using microneedle array electrodes and polymer-based strain sensors","volume":"11","author":"Lin","year":"2025","journal-title":"Sens. Actuator Rep."},{"key":"jneae54d0bib12","doi-asserted-by":"publisher","DOI":"10.1016\/j.bspc.2022.104298","type":"journal-article","article-title":"Decoding silent speech from high-density surface electromyographic data using transformer","volume":"80","author":"Song","year":"2023","journal-title":"Biomed. Signal Process. Control"},{"key":"jneae54d0bib13","doi-asserted-by":"publisher","DOI":"10.1038\/srep16743","type":"journal-article","article-title":"Unobtrusive ambulatory EEG using a smartphone and flexible printed electrodes around the ear","volume":"5","author":"Debener","year":"2015","journal-title":"Sci. Rep."},{"key":"jneae54d0bib14","article-title":"Wireless silent speech interface using multi-channel textile EMG sensors integrated into headphones","author":"Tang","year":"2025","type":"preprint"},{"key":"jneae54d0bib15","article-title":"Scaling law in neural data: Non-invasive speech decoding with 175 h of EEG data","author":"Sato","year":"2024","type":"preprint"},{"key":"jneae54d0bib16","article-title":"Scaling laws for decoding images from brain activity","author":"Banville","year":"2025","type":"preprint"},{"key":"jneae54d0bib17","article-title":"Brain-to-text decoding: a non-invasive approach via typing","author":"L\u00e9vy","year":"2025","type":"preprint"},{"key":"jneae54d0bib18","doi-asserted-by":"publisher","DOI":"10.1038\/s41467-025-65499-0","type":"journal-article","article-title":"Towards decoding individual words from non-invasive brain recordings","volume":"16","author":"d\u2019Ascoli","year":"2025","journal-title":"Nat. Commun."},{"key":"jneae54d0bib19","doi-asserted-by":"publisher","first-page":"6744","DOI":"10.3390\/s21206744","type":"journal-article","article-title":"Silent EEG-speech recognition using convolutional and recurrent neural network with 85% accuracy of 9 words classification","volume":"21","author":"Vorontsova","year":"2021","journal-title":"Sensors"},{"key":"jneae54d0bib20","doi-asserted-by":"publisher","DOI":"10.14569\/IJACSA.2026.0170136","type":"journal-article","article-title":"A systematic review of EEG-based imagined speech decoding","volume":"17","author":"Jin","year":"2025","journal-title":"Appl. Soft Comput."},{"key":"jneae54d0bib21","doi-asserted-by":"publisher","first-page":"2625","DOI":"10.1044\/2020_JSLHR-19-00309","type":"journal-article","article-title":"Tongue- and jaw-specific articulatory changes and their acoustic consequences in talkers with dysarthria due to amyotrophic lateral sclerosis: effects of loud, clear and slow speech","volume":"63","author":"Mefferd","year":"2020","journal-title":"J. Speech Lang. Hear. Res."},{"key":"jneae54d0bib22","doi-asserted-by":"publisher","first-page":"725","DOI":"10.1080\/02699206.2017.1318173","type":"journal-article","article-title":"An investigation of clear speech effects on articulatory kinematics in talkers with ALS","volume":"31","author":"Kuruvilla-Dugdale","year":"2017","journal-title":"Clin. Linguist. Phon."},{"key":"jneae54d0bib23","doi-asserted-by":"publisher","first-page":"2257","DOI":"10.1109\/TASLP.2017.2752365","type":"journal-article","article-title":"Biosignal-based spoken communication: a survey","volume":"25","author":"Schultz","year":"2017","journal-title":"IEEE\/ACM Trans. Audio, Speech Language Process."},{"key":"jneae54d0bib24","doi-asserted-by":"publisher","first-page":"648","DOI":"10.11477\/mf.1401204368","type":"journal-article","article-title":"Basic issue on severe motor and intellectual disabilities","volume":"35","author":"Oshima","year":"1971","journal-title":"Nippon Koshu Eisei Zasshi"},{"key":"jneae54d0bib25","doi-asserted-by":"publisher","first-page":"77","DOI":"10.2188\/jea.JE20090024","type":"journal-article","article-title":"Survival prognosis of Japanese with severe motor and intellectual disabilities living in public and private institutions between 1961 and 2003","volume":"20","author":"Hanaoka","year":"2010","journal-title":"J. Epidemiol."},{"key":"jneae54d0bib26","doi-asserted-by":"publisher","first-page":"31","DOI":"10.1002\/9781405166225.ch2","type":"book","article-title":"Mora and syllable","author":"Kubozono","year":"2017"},{"key":"jneae54d0bib27","doi-asserted-by":"publisher","DOI":"10.1016\/j.ohx.2022.e00357","type":"journal-article","article-title":"A simplified design of a cEEGrid ear-electrode adapter for the OpenBCI biosensing platform","volume":"12","author":"Knierim","year":"2022","journal-title":"HardwareX"},{"key":"jneae54d0bib28","doi-asserted-by":"publisher","first-page":"3996","DOI":"10.1109\/JSEN.2019.2962874","type":"journal-article","article-title":"Consumer grade EEG measuring sensors as research tools: a review","volume":"20","author":"Sawangjai","year":"2019","journal-title":"IEEE Sens. J."},{"key":"jneae54d0bib29","doi-asserted-by":"publisher","first-page":"5603","DOI":"10.21437\/Interspeech.2025-1183","type":"conference-proceedings","article-title":"A silent speech decoding system from EEG and EMG with heterogenous electrode configurations","author":"Inoue","year":"2025"},{"key":"jneae54d0bib30","first-page":"28492","type":"conference-proceedings","article-title":"Robust speech recognition via large-scale weak supervision","author":"Radford","year":"2023"},{"key":"jneae54d0bib31","article-title":"MeCab: yet another part-of-speech and morphological analyzer","author":"Kudo","year":"2005","type":"other"},{"key":"jneae54d0bib32","first-page":"230","type":"conference-proceedings","article-title":"Applying conditional random fields to Japanese morphological analysis","author":"Kudo","year":"2004"},{"key":"jneae54d0bib33","article-title":"Silero VAD: pre-trained enterprise-grade voice activity detector (VAD), number detector and language classifier","author":"Team","year":"2024","type":"other"},{"key":"jneae54d0bib34","doi-asserted-by":"publisher","DOI":"10.1088\/1741-2552\/abda0b","type":"journal-article","article-title":"Generalized neural decoders for transfer learning across participants and recording modalities","volume":"18","author":"Peterson","year":"2021","journal-title":"J. Neural Eng."},{"key":"jneae54d0bib35","doi-asserted-by":"crossref","DOI":"10.21437\/Interspeech.2020-3015","type":"preprint","article-title":"Conformer: convolution-augmented transformer for speech recognition","author":"Gulati","year":"2020"},{"key":"jneae54d0bib36","doi-asserted-by":"publisher","first-page":"3","DOI":"10.1007\/978-3-030-01261-8_1","type":"conference-proceedings","article-title":"Group normalization","author":"Wu","year":"2018"},{"key":"jneae54d0bib37","article-title":"Gaussian error linear units (gelus)","author":"Hendrycks","year":"2016","type":"preprint"},{"key":"jneae54d0bib38","type":"conference-proceedings","article-title":"Attention is all you need","volume":"vol 30","author":"Vaswani","year":"2017"},{"key":"jneae54d0bib39","first-page":"12449","type":"conference-proceedings","article-title":"wav2vec 2.0: a framework for self-supervised learning of speech representations","volume":"vol 33","author":"Baevski","year":"2020"},{"key":"jneae54d0bib40","article-title":"Layer normalization","author":"Ba","year":"2016","type":"preprint"},{"key":"jneae54d0bib41","doi-asserted-by":"publisher","first-page":"369","DOI":"10.1145\/1143844.1143891","type":"conference-proceedings","article-title":"Connectionist temporal classification: labelling unsegmented sequence data with recurrent neural networks","author":"Graves","year":"2006"},{"key":"jneae54d0bib42","article-title":"Decoupled weight decay regularization","author":"Loshchilov","year":"2017","type":"preprint"},{"key":"jneae54d0bib43","doi-asserted-by":"publisher","DOI":"10.52202\/079017-1723","type":"conference-proceedings","article-title":"Brainbits: how much of the brain are generative reconstruction methods using?","author":"Mayo","year":"2025"},{"key":"jneae54d0bib44","doi-asserted-by":"crossref","DOI":"10.18653\/v1\/2020.emnlp-main.445","type":"preprint","article-title":"Digital voicing of silent speech","author":"Gaddy","year":"2020"},{"key":"jneae54d0bib45","doi-asserted-by":"publisher","DOI":"10.3389\/fnins.2023.1122661","type":"journal-article","article-title":"Discrepancy between inter- and intra-subject variability in EEG-based motor imagery brain-computer interface: evidence from multiple perspectives","volume":"17","author":"Huang","year":"2023","journal-title":"Front. Neurosci."},{"key":"jneae54d0bib46","doi-asserted-by":"publisher","first-page":"243","DOI":"10.5582\/bst.2025.01109","type":"journal-article","article-title":"Integrative neurorehabilitation using brain-computer interface: From motor function to mental health after stroke","volume":"19","author":"Ma","year":"2025","journal-title":"BioSci. Trends"},{"key":"jneae54d0bib47","doi-asserted-by":"publisher","first-page":"87","DOI":"10.3389\/fncom.2019.00087","type":"journal-article","article-title":"Intra- and inter-subject variability in EEG-based sensorimotor brain computer interface: a review","volume":"13","author":"Saha","year":"2020","journal-title":"Front. Comput. Neurosci."},{"key":"jneae54d0bib48","doi-asserted-by":"publisher","first-page":"2282","DOI":"10.1109\/TCDS.2023.3316701","type":"journal-article","article-title":"EMG-based cross-subject silent speech recognition using conditional domain adversarial network","volume":"15","author":"Zhang","year":"2023","journal-title":"IEEE Trans. Cogn. Dev. Syst."},{"key":"jneae54d0bib49","doi-asserted-by":"publisher","DOI":"10.1088\/1741-2552\/ad152f","type":"journal-article","article-title":"Transferring a deep learning model from healthy subjects to stroke patients in a motor imagery brain\u2013computer interface","volume":"21","author":"Nagarajan","year":"2024","journal-title":"J. Neural Eng."},{"key":"jneae54d0bib50","doi-asserted-by":"publisher","first-page":"1097","DOI":"10.1038\/s42256-023-00714-5","type":"journal-article","article-title":"Decoding speech perception from non-invasive brain recordings","volume":"5","author":"D\u00e9fossez","year":"2023","journal-title":"Nat. Mach. Intell."},{"key":"jneae54d0bib51","doi-asserted-by":"publisher","DOI":"10.1016\/j.compbiomed.2022.106420","type":"journal-article","article-title":"MI-DABAN: a dual-attention-based adversarial network for motor imagery classification","volume":"152","author":"Li","year":"2023","journal-title":"Comput. Biol. Med."},{"key":"jneae54d0bib52","doi-asserted-by":"publisher","first-page":"30","DOI":"10.1007\/978-3-030-36708-4_3","type":"conference-proceedings","article-title":"Reducing the subject variability of EEG signals with adversarial domain generalization","author":"Ma","year":"2019"},{"key":"jneae54d0bib53","first-page":"1779","type":"journal-article","article-title":"Club: a contrastive log-ratio upper bound of mutual information","author":"Cheng","year":"2020"}],"container-title":["Journal of Neural Engineering"],"original-title":[],"link":[{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/1741-2552\/ae54d0","content-type":"text\/html","content-version":"am","intended-application":"text-mining"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/1741-2552\/ae54d0\/pdf","content-type":"application\/pdf","content-version":"am","intended-application":"text-mining"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/1741-2552\/ae54d0","content-type":"text\/html","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/1741-2552\/ae54d0\/pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/1741-2552\/ae54d0\/pdf","content-type":"application\/pdf","content-version":"am","intended-application":"syndication"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/1741-2552\/ae54d0\/pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"syndication"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/1741-2552\/ae54d0\/pdf","content-type":"application\/pdf","content-version":"am","intended-application":"similarity-checking"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/1741-2552\/ae54d0\/pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,4,7]],"date-time":"2026-04-07T07:03:31Z","timestamp":1775545411000},"score":1,"resource":{"primary":{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/1741-2552\/ae54d0"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,4,1]]},"references-count":53,"journal-issue":{"issue":"2","published-online":{"date-parts":[[2026,4,7]]},"published-print":{"date-parts":[[2026,4,1]]}},"URL":"https:\/\/doi.org\/10.1088\/1741-2552\/ae54d0","relation":{},"ISSN":["1741-2560","1741-2552"],"issn-type":[{"value":"1741-2560","type":"print"},{"value":"1741-2552","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,4,1]]},"assertion":[{"value":"Large-scale training data enhances silent speech decoding with around-ear EEG","name":"article_title","label":"Article Title"},{"value":"Journal of Neural Engineering","name":"journal_title","label":"Journal Title"},{"value":"paper","name":"article_type","label":"Article Type"},{"value":"\u00a9 2026 Araya Inc. Published by IOP Publishing Ltd","name":"copyright_information","label":"Copyright Information"},{"value":"2025-12-24","name":"date_received","label":"Date Received","group":{"name":"publication_dates","label":"Publication dates"}},{"value":"2026-03-19","name":"date_accepted","label":"Date Accepted","group":{"name":"publication_dates","label":"Publication dates"}},{"value":"2026-04-07","name":"date_epub","label":"Online publication date","group":{"name":"publication_dates","label":"Publication dates"}}]}}