{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,8,4]],"date-time":"2026-08-04T05:47:30Z","timestamp":1785822450121,"version":"3.56.0"},"reference-count":46,"publisher":"Frontiers Media SA","license":[{"start":{"date-parts":[[2026,8,4]],"date-time":"2026-08-04T00:00:00Z","timestamp":1785801600000},"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. Hum. Neurosci."],"abstract":"<jats:sec>\n                    <jats:title>Introduction<\/jats:title>\n                    <jats:p>Pediatric speech sound disorders (SSDs) affect many young children and are commonly assessed through auditory-perceptual judgments and IPA transcription, which are limited by listener bias, variable interrater reliability, and difficulty attributing deviations at the phoneme level. We present a developmentally informed framework for pediatric vocal biomarker foundations that prioritizes age-aware, phoneme-resolved interpretability over utterance-level accuracy alone.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Methods<\/jats:title>\n                    <jats:p>Using a CAAP-derived subset of the SEED corpus (27\u201394 months) with clinically specified phoneme targets and SSD labels, we construct age-stratified phoneme profiles and quantify error structure via phoneme error rate (PER) and age-banded confusion signatures. We operationalize co-articulation as transition dynamics from MFCC trajectories and their first and second derivatives, reflecting the velocity and acceleration of spectral change across adjacent segments. We learn compact acoustic representations with a variational autoencoder (VAE) and model temporal evolution with BiLSTMs, including attention, to characterize disorder-relevant instability in latent trajectories. For phoneme transcription, we train BiLSTM-CTC sequence models on clinically elicited speech and evaluate disorder classification and phoneme substitution patterns within age groups, then stress-test generalization on ECSC \"Frog Story\" narratives from TalkBank\/CHILDES.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Results<\/jats:title>\n                    <jats:p>Co-articulation transition magnitudes varied systematically by age and clinical group. Attention improved temporal classification specifically in younger children, and PER dropped markedly between ages 3\u20134 and 4\u20135. On long-form naturalistic narratives, CTC training and inference remained numerically stable, but greedy decoding produced degenerate output dominated by blank and repetition tokens.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Discussion<\/jats:title>\n                    <jats:p>Together, these results support an age-aware approach to pediatric phoneme analytics, in which articulatory coordination, not just phoneme identity, carries developmentally and clinically relevant information; and phoneme-level performance should be interpreted relative to developmental stage rather than a single fixed benchmark. The decoding failures on naturalistic speech indicate a bottleneck specific to decoding rather than to the underlying representations, motivating constrained, hierarchical, or duration-aware decoding as a tractable next step. These findings position age-stratified, phoneme-resolved analysis as a foundation for interpretable and scalable pediatric screening and biomarker development.<\/jats:p>\n                  <\/jats:sec>","DOI":"10.3389\/fnhum.2026.1802703","type":"journal-article","created":{"date-parts":[[2026,8,4]],"date-time":"2026-08-04T05:39:08Z","timestamp":1785821948000},"update-policy":"https:\/\/doi.org\/10.3389\/crossmark-policy","source":"Crossref","is-referenced-by-count":0,"title":["Foundations for pediatric vocal biomarkers: age-aware phoneme recognition and latent-space error analysis"],"prefix":"10.3389","volume":"20","author":[{"given":"Chethana","family":"Saligram","sequence":"first","affiliation":[{"name":"Roxelyn and Richard Pepper Department of Communication Disorders, Northwestern University","place":["Evanston, IL, United States"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Vishal","family":"Shrivastava","sequence":"additional","affiliation":[{"name":"Roxelyn and Richard Pepper Department of Communication Disorders, Northwestern University","place":["Evanston, IL, United States"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Marisha","family":"Speights","sequence":"additional","affiliation":[{"name":"Roxelyn and Richard Pepper Department of Communication Disorders, Northwestern University","place":["Evanston, IL, United States"]}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"1965","published-online":{"date-parts":[[2026,8,4]]},"reference":[{"key":"B1","unstructured":"Speech sound disorders: articulation and phonology\n          \n          2025"},{"key":"B2","first-page":"12449","article-title":"\u201cwav2vec 2.0: a framework for self-supervised learning of speech representations,\u201d","author":"Baevski","year":"2020","journal-title":"Advances in neural information processing systems"},{"key":"B3","doi-asserted-by":"publisher","first-page":"442","DOI":"10.1044\/2025_LSHSS-24-00085","article-title":"Equipping speech-language clinicians for the critical appraisal of an artificial intelligence-driven, evidence-based future","volume":"56","author":"Benway","year":"2025","journal-title":"Lang. Speech Hear. Serv. Schools"},{"key":"B4","doi-asserted-by":"publisher","first-page":"208","DOI":"10.1038\/s41746-024-01199-1","article-title":"Responsible development of clinical speech AI: bridging the gap between clinical research and technology","volume":"7","author":"Berisha","year":"2024","journal-title":"NPJ Digital Med"},{"key":"B5","volume-title":"Relating Events in Narrative: A Crosslinguistic Developmental Study","author":"Berman","year":"1994"},{"key":"B6","doi-asserted-by":"publisher","first-page":"4419","DOI":"10.3390\/app12094419","article-title":"Automatic speech recognition (asr) systems for children: a systematic literature review","volume":"12","author":"Bhardwaj","year":"2022","journal-title":"Appl. Sci"},{"key":"B7","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-319-44781-0_8","article-title":"Layer-wise relevance propagation for neural networks with local renormalization layers","author":"Binder","year":"2016","journal-title":"arXiv, abs\/1604.00825"},{"key":"B8","doi-asserted-by":"publisher","first-page":"189","DOI":"10.1007\/s40474-014-0017-3","article-title":"Differential diagnosis of pediatric speech sound disorder","volume":"1","author":"Dodd","year":"2014","journal-title":"Curr. Dev. Disor. Rep"},{"key":"B9","doi-asserted-by":"publisher","DOI":"10.1145\/1640377.1640384","article-title":"\u201cA review of asr technologies for children's speech,\u201d","author":"Gerosa","year":"2009","journal-title":"Proceedings of the 2nd workshop on child, computer and interaction"},{"key":"B10","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1561\/2200000089","article-title":"Dynamical variational autoencoders: a comprehensive review","volume":"15","author":"Girin","year":"2022","journal-title":"Found. Trends Mach. Learn"},{"key":"B11","doi-asserted-by":"publisher","DOI":"10.1145\/1143844.1143891","article-title":"\u201cConnectionist temporal classification: labelling unsegmented sequence data with recurrent neural networks,\u201d","author":"Graves","year":"2006","journal-title":"Proceedings of the 23rd international conference on machine learning"},{"key":"B12","article-title":"\u201cUnsupervised learning of disentangled and interpretable representations from sequential data,\u201d","author":"Hsu","year":"2017","journal-title":"Advances in neural information processing systems"},{"key":"B13","doi-asserted-by":"publisher","first-page":"3","DOI":"10.1017\/S030500099900402X","article-title":"The phonological acquisition of Putonghua (Modern Standard Chinese)","volume":"27","author":"Hua","year":"2000","journal-title":"J. Child Lang"},{"key":"B14","doi-asserted-by":"publisher","year":"1999","journal-title":"Handbook of the International Phonetic Association: A guide to the use of the International Phonetic Alphabet","DOI":"10.1017\/9780511807954"},{"key":"B15","doi-asserted-by":"publisher","first-page":"88","DOI":"10.1017\/S0305000920000197","article-title":"Clause-initial and usage in a cross-sectional and longitudinal corpus of school-age children's narratives","volume":"48","author":"Kallay","year":"2020","journal-title":"J. Child Lang"},{"key":"B16","doi-asserted-by":"publisher","first-page":"1214","DOI":"10.1080\/23273798.2019.1589541","article-title":"Articulating: the neural mechanisms of speech production","volume":"34","author":"Kearney","year":"2019","journal-title":"Lang. Cogn. Neurosci"},{"key":"B17","first-page":"90","article-title":"The biology of phonological development","volume":"65","author":"Kent","year":"1992","journal-title":"Phonol. Dev"},{"key":"B18","doi-asserted-by":"publisher","first-page":"7","DOI":"10.1044\/1058-0360.0503.07","article-title":"Hearing and believing: some limits to the auditory-perceptual assessment of speech and voice disorders","volume":"5","author":"Kent","year":"1996","journal-title":"Am. J. Speech-Lang. Pathol"},{"key":"B19","article-title":"Auto-encoding variational bayes","author":"Kingma","year":"2013","journal-title":"arXiv preprint arXiv:1312.6114"},{"key":"B20","doi-asserted-by":"publisher","first-page":"844529","DOI":"10.3389\/fnhum.2022.844529","article-title":"On the emergence of phonological knowledge and on motor planning and motor programming in a developmental model of speech production","volume":"16","author":"Kr\u00f6ger","year":"2022","journal-title":"Front. Hum. Neurosci"},{"key":"B21","doi-asserted-by":"publisher","DOI":"10.21437\/Interspeech.2018-1568","article-title":"\u201cVariational autoencoders for learning latent representations of speech emotion: a preliminary study,\u201d","author":"Latif","year":"2018","journal-title":"Interspeech 2018"},{"key":"B22","doi-asserted-by":"publisher","first-page":"1455","DOI":"10.1121\/1.426686","article-title":"Acoustics of children's speech: developmental changes of temporal and spectral parameters","volume":"105","author":"Lee","year":"1999","journal-title":"J. Acoust. Soc. Am"},{"key":"B23","doi-asserted-by":"publisher","first-page":"339","DOI":"10.1044\/jshd.4804.339","article-title":"Clinical phonology: the explanation and treatment of speech sound disorders","volume":"48","author":"Locke","year":"1983","journal-title":"J. Speech Hear. Disor"},{"key":"B24","doi-asserted-by":"publisher","first-page":"285","DOI":"10.5772\/17600","article-title":"Phone recognition on the timit database","volume":"1","author":"Lopes","year":"2011","journal-title":"Speech Technol. Book"},{"key":"B25","doi-asserted-by":"publisher","first-page":"809722","DOI":"10.3389\/fpsyg.2022.809722","article-title":"The phonological development of mandarin voiceless affricates in three- to five-year-old children","volume":"13","author":"Ma","year":"2022","journal-title":"Front. Psychol"},{"key":"B26","doi-asserted-by":"publisher","first-page":"153","DOI":"10.1111\/1467-8624.00129","article-title":"The motor core of speech: a comparison of serial organization patterns in infants and languages","volume":"71","author":"MacNeilage","year":"2000","journal-title":"Child Dev"},{"key":"B27","doi-asserted-by":"publisher","first-page":"100913","DOI":"10.1016\/j.measen.2023.100913","article-title":"Dysarthria detection using convolution neural network","volume":"30","author":"Mahendran","year":"2023","journal-title":"Measurement"},{"key":"B28","doi-asserted-by":"publisher","first-page":"245","DOI":"10.3109\/02699200903532524","article-title":"Deconstructing phonetic transcription: covert contrast, perceptual bias, and an extraterrestrial view of vox humana","volume":"24","author":"Munson","year":"2010","journal-title":"Clin. Ling. Phonet"},{"key":"B29","doi-asserted-by":"publisher","first-page":"2998","DOI":"10.3389\/fpsyg.2019.02998","article-title":"Speech sound disorders in children: an articulatory phonology perspective","volume":"10","author":"Namasivayam","year":"2020","journal-title":"Front. Psychol"},{"key":"B30","doi-asserted-by":"publisher","first-page":"1700505","DOI":"10.3389\/fnhum.2025.1700505","article-title":"Reevaluating the classification of pediatric speech sound disorders: a ground truthing perspective","volume":"19","author":"Namasivayam","year":"2025","journal-title":"Front. Hum. Neurosci"},{"key":"B31","year":"2025","journal-title":"Quick Statistics About Voice, Speech, Language"},{"key":"B32","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1109\/ICASSP49357.2023.10096149","article-title":"\u201cComparative layer-wise analysis of self-supervised speech models,\u201d","author":"Pasad","year":"2023","journal-title":"ICASSP 2023\u20132023 IEEE international conference on acoustics, speech and signal processing (ICASSP)"},{"key":"B33","doi-asserted-by":"publisher","DOI":"10.1101\/2025.10.23.25338518","article-title":"Consensus-based definitions for vocal biomarkers: The international vocal initiative","author":"Pizzimenti","year":"2025","journal-title":"medRxiv, 2025.10.23.25338518"},{"key":"B34","doi-asserted-by":"publisher","first-page":"603","DOI":"10.1109\/TSA.2003.818026","article-title":"Robust recognition of children's speech","volume":"11","author":"Potamianos","year":"2004","journal-title":"IEEE Trans. Speech Audio Proc"},{"key":"B35","doi-asserted-by":"publisher","first-page":"206","DOI":"10.1038\/s42256-019-0048-x","article-title":"Stop explaining black box machine learning models for high stakes decisions and use interpretable models instead","volume":"1","author":"Rudin","year":"2019","journal-title":"Nat. Mach. Intell"},{"key":"B36","author":"Secord","year":"2002","journal-title":"CAAP: Clinical Assessment of Articulation and Phonology"},{"key":"B37","doi-asserted-by":"publisher","first-page":"400","DOI":"10.1109\/JSTSP.2019.2959393","article-title":"The automatic detection of speech disorders in children: challenges, opportunities, and preliminary results","volume":"14","author":"Shahin","year":"2019","journal-title":"IEEE J. Sel. Top. Signal Process"},{"key":"B38","doi-asserted-by":"publisher","first-page":"101077","DOI":"10.1016\/j.csl.2020.101077","article-title":"Transfer learning from adult to children for speech recognition: evaluation, analysis and recommendations","volume":"63","author":"Shivakumar","year":"2020","journal-title":"Comput. Speech Lang"},{"key":"B39","doi-asserted-by":"publisher","first-page":"825","DOI":"10.3109\/02699206.2010.503007","article-title":"Perceptual and acoustic reliability estimates for the speech disorders classification system (sdcs)","volume":"24","author":"Shriberg","year":"2010","journal-title":"Clin. Ling. Phonet"},{"key":"B40","doi-asserted-by":"publisher","first-page":"878","DOI":"10.1080\/02699206.2020.1743761","article-title":"Speech exemplar and evaluation database (seed) for clinical training in articulatory phonetics and speech science","volume":"34","author":"Speights Atkins","year":"2020","journal-title":"Clin. Ling. Phonet"},{"key":"B41","doi-asserted-by":"publisher","first-page":"975","DOI":"10.1016\/j.jvoice.2022.03.021","article-title":"The effect of the mfcc frame length in automatic voice pathology detection","volume":"38","author":"Tirronen","year":"2022","journal-title":"J. Voice"},{"key":"B42","doi-asserted-by":"publisher","first-page":"338","DOI":"10.1121\/1.1835958","article-title":"Development of vocal tract length during early childhood: a magnetic resonance imaging study","volume":"117","author":"Vorperian","year":"2005","journal-title":"J. Acoust. Soc. Am"},{"key":"B43","doi-asserted-by":"publisher","first-page":"1637","DOI":"10.1044\/2025_PERSP-25-00003","article-title":"Evaluating model interpretability in speech-based clinical artificial intelligence systems","volume":"10","author":"Xu","year":"2025","journal-title":"Persp. ASHA Special Interest Groups"},{"key":"B44","doi-asserted-by":"publisher","DOI":"10.21437\/Interspeech.2018-2297","article-title":"\u201cOn the difficulties of automatic speech recognition for kindergarten-aged children,\u201d","author":"Yeung","year":"2018","journal-title":"Interspeech 2018"},{"key":"B45","doi-asserted-by":"publisher","first-page":"1091","DOI":"10.1109\/TPAMI.2007.1078","article-title":"A normalized levenshtein distance metric","volume":"29","author":"Yujian","year":"2007","journal-title":"IEEE Trans. Pattern Anal. Mach. Intell"},{"key":"B46","doi-asserted-by":"publisher","first-page":"726","DOI":"10.1109\/TETCI.2021.3100641","article-title":"A survey on neural network interpretability","volume":"5","author":"Zhang","year":"2021","journal-title":"IEEE Trans. Emer. Topics Comput. Intell"}],"container-title":["Frontiers in Human Neuroscience"],"original-title":[],"link":[{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/fnhum.2026.1802703\/full","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,8,4]],"date-time":"2026-08-04T05:39:10Z","timestamp":1785821950000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/fnhum.2026.1802703\/full"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,8,4]]},"references-count":46,"alternative-id":["10.3389\/fnhum.2026.1802703"],"URL":"https:\/\/doi.org\/10.3389\/fnhum.2026.1802703","relation":{},"ISSN":["1662-5161"],"issn-type":[{"value":"1662-5161","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,8,4]]},"article-number":"1802703"}}