{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"institution":[{"name":"medRxiv"}],"indexed":{"date-parts":[[2026,1,16]],"date-time":"2026-01-16T12:02:22Z","timestamp":1768564942588,"version":"3.49.0"},"posted":{"date-parts":[[2023,4,4]]},"group-title":"Neurology","reference-count":36,"publisher":"openRxiv","content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"accepted":{"date-parts":[[2023,4,4]]},"abstract":"<jats:title>Abstract<\/jats:title>\n                <jats:sec>\n                  <jats:title>Objective<\/jats:title>\n                  <jats:p>Objective, sensitive, and meaningful disease assessments are critical to support clinical trials and clinical care. Speech changes are one of the earliest and most evident manifestations of cerebellar ataxias. The purpose of this work is to develop models that can accurately identify and quantify these abnormalities.<\/jats:p>\n                <\/jats:sec>\n                <jats:sec>\n                  <jats:title>Methods<\/jats:title>\n                  <jats:p>\n                    We use deep learning models such as\n                    <jats:italic>ResNet 18<\/jats:italic>\n                    , that take the time and frequency partial derivatives of the log-mel spectrogram representations of speech as input, to learn representations that capture the motor speech phenotype of cerebellar ataxia. We train classification models to separate patients with ataxia from healthy controls as well as regression models to estimate disease severity.\n                  <\/jats:p>\n                <\/jats:sec>\n                <jats:sec>\n                  <jats:title>Results<\/jats:title>\n                  <jats:p>Our model was able to accurately distinguish healthy controls from individuals with ataxia, including ataxia participants with no detectable clinical deficits in speech. Furthermore the regression models produced accurate estimates of disease severity, were able to measure subclinical signs of ataxia, and captured disease progression over time in individuals with ataxia.<\/jats:p>\n                <\/jats:sec>\n                <jats:sec>\n                  <jats:title>Conclusion<\/jats:title>\n                  <jats:p>Deep learning models, trained on time and frequency partial derivatives of the speech signal, can detect sub-clinical speech changes in ataxias and sensitively measure disease change over time.<\/jats:p>\n                <\/jats:sec>\n                <jats:sec>\n                  <jats:title>Significance<\/jats:title>\n                  <jats:p>Such models have the potential to assist with early detection of ataxia and to provide sensitive and low-burden assessment tools in support of clinical trials and neurological care.<\/jats:p>\n                <\/jats:sec>","DOI":"10.1101\/2023.04.03.23288094","type":"posted-content","created":{"date-parts":[[2023,4,4]],"date-time":"2023-04-04T18:10:11Z","timestamp":1680631811000},"source":"Crossref","is-referenced-by-count":1,"title":["Sensitive quantification of cerebellar speech abnormalities using deep learning models"],"prefix":"10.64898","author":[{"given":"Kyriakos","family":"Vattis","sequence":"first","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Anna C.","family":"Luddy","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Jessey S.","family":"Ouillon","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Nicole M.","family":"Eklund","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Christopher D.","family":"Stephen","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Jeremy D.","family":"Schmahmann","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Adonay S.","family":"Nunes","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-8741-0621","authenticated-orcid":false,"given":"Anoopum S.","family":"Gupta","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"54368","reference":[{"key":"2023040612500572000_2023.04.03.23288094v1.1","doi-asserted-by":"publisher","DOI":"10.1159\/000358801"},{"key":"2023040612500572000_2023.04.03.23288094v1.2","doi-asserted-by":"publisher","DOI":"10.1212\/01.wnl.0000219042.60538.92"},{"key":"2023040612500572000_2023.04.03.23288094v1.3","doi-asserted-by":"publisher","DOI":"10.1016\/S0022-510X(96)00231-6"},{"key":"2023040612500572000_2023.04.03.23288094v1.4","doi-asserted-by":"publisher","DOI":"10.1002\/mds.22681"},{"key":"2023040612500572000_2023.04.03.23288094v1.5","doi-asserted-by":"crossref","unstructured":"A. S. Gupta , \u201cDigital phenotyping in clinical neurology,\u201d in Seminars in neurology. Thieme Medical Publishers, Inc., 2022.","DOI":"10.1055\/s-0041-1741495"},{"key":"2023040612500572000_2023.04.03.23288094v1.6","article-title":"Assessing Cerebellar Disorders With Wearable Inertial Sensor Data Using Time-Frequency and Autoregressive Hidden Markov Model Approaches","year":"2021","journal-title":"arXiv e-prints"},{"issue":"9","key":"2023040612500572000_2023.04.03.23288094v1.7","doi-asserted-by":"crossref","first-page":"e1199","DOI":"10.1212\/WNL.0000000000010176","article-title":"Real-life gait assessment in degenerative cerebellar ataxia","volume":"95","year":"2020","journal-title":"Neurology"},{"key":"2023040612500572000_2023.04.03.23288094v1.8","doi-asserted-by":"crossref","unstructured":"A. S. Gupta , A. C. Luddy , N. C. Khan , S. Reiling , and J. K. Thornton , \u201cReal-life wrist movement patterns capture motor impairment in individuals with ataxia-telangiectasia,\u201d The Cerebellum, pp. 1\u201311, 2022.","DOI":"10.1007\/s12311-022-01385-5"},{"key":"2023040612500572000_2023.04.03.23288094v1.9","doi-asserted-by":"crossref","unstructured":"J. Lee , B. Oubre , J.-F. Daneault , C. D. Stephen , J. D. Schmahmann , A. S. Gupta , and S. I. Lee , \u201cAnalysis of gait sub-movements to estimate ataxia severity using ankle inertial data,\u201d IEEE Transactions on Biomedical Engineering, 2022.","DOI":"10.1109\/TBME.2022.3142504"},{"key":"2023040612500572000_2023.04.03.23288094v1.10","doi-asserted-by":"crossref","unstructured":"N. C. Khan , V. Pandey , K. Z. Gajos , and A. S. Gupta , \u201cFree-living motor activity monitoring in ataxia-telangiectasia,\u201d The Cerebellum, pp. 1\u201312, 2021.","DOI":"10.1007\/s12311-021-01306-y"},{"issue":"6","key":"2023040612500572000_2023.04.03.23288094v1.11","doi-asserted-by":"crossref","first-page":"811","DOI":"10.1007\/s12311-021-01247-6","article-title":"Decomposition of reaching movements enables detection and measurement of ataxia","volume":"20","year":"2021","journal-title":"The Cerebellum"},{"key":"2023040612500572000_2023.04.03.23288094v1.12","doi-asserted-by":"crossref","first-page":"354","DOI":"10.1002\/mds.27915","article-title":"Computer mouse use captures ataxia and parkinsonism, enabling accurate measurement and detection","volume":"35","year":"2020","journal-title":"Movement Disorders"},{"key":"2023040612500572000_2023.04.03.23288094v1.13","doi-asserted-by":"publisher","DOI":"10.1016\/j.clinph.2008.12.040"},{"issue":"1","key":"2023040612500572000_2023.04.03.23288094v1.14","first-page":"1","article-title":"Accurate detection of cerebellar smooth pursuit eye movement abnormalities via mobile phone video and machine learning","volume":"10","year":"2020","journal-title":"Scientific reports"},{"key":"2023040612500572000_2023.04.03.23288094v1.15","doi-asserted-by":"crossref","unstructured":"M. Asgari and I. Shafran , \u201cExtracting cues from speech for predicting severity of parkinson\u2019s disease,\u201d in 2010 IEEE International Workshop on Machine Learning for Signal Processing. IEEE, 2010, pp. 462\u2013467.","DOI":"10.1109\/MLSP.2010.5589118"},{"key":"2023040612500572000_2023.04.03.23288094v1.16","doi-asserted-by":"crossref","unstructured":"H. Hazan , D. Hilu , L. Manevitz , L. Ramig , and S. Sapir , \u201cEarly diagnosis of parkinson\u2019s disease via machine learning on speech data, in 2012 ieee 27th convention of electrical and electronics engineers in israel,\u201d Google Scholar Google Scholar Cross Ref Cross Ref, 2012.","DOI":"10.1109\/EEEI.2012.6377065"},{"key":"2023040612500572000_2023.04.03.23288094v1.17","doi-asserted-by":"publisher","DOI":"10.1109\/TBME.2012.2183367"},{"issue":"1","key":"2023040612500572000_2023.04.03.23288094v1.18","doi-asserted-by":"crossref","first-page":"263","DOI":"10.1016\/j.eswa.2012.07.014","article-title":"An efficient diagnosis system for detection of parkinson\u2019s disease using fuzzy k-nearest neighbor approach","volume":"40","year":"2013","journal-title":"Expert systems with applications"},{"issue":"10","key":"2023040612500572000_2023.04.03.23288094v1.19","doi-asserted-by":"crossref","first-page":"12 470","DOI":"10.1016\/j.eswa.2011.04.028","article-title":"A parallel neural network approach to prediction of parkinson\u2019s disease","volume":"38","year":"2011","journal-title":"Expert systems with applications"},{"key":"2023040612500572000_2023.04.03.23288094v1.20","doi-asserted-by":"crossref","unstructured":"D. Jain , A. K. Mishra , and S. K. Das , \u201cMachine learning based automatic prediction of parkinson\u2019s disease using speech features,\u201d in Proceedings of International Conference on Artificial Intelligence and Applications. Springer, 2021, pp. 351\u2013362.","DOI":"10.1007\/978-981-15-4992-2_33"},{"key":"2023040612500572000_2023.04.03.23288094v1.21","doi-asserted-by":"crossref","first-page":"10 239","DOI":"10.1109\/ACCESS.2021.3051432","article-title":"A deep learning based method for parkinson\u2019s disease detection using dynamic features of speech","volume":"9","year":"2021","journal-title":"IEEE Access"},{"issue":"1","key":"2023040612500572000_2023.04.03.23288094v1.22","first-page":"1","article-title":"Early detection and tracking of bulbar changes in als via frequent and remote speech analysis","volume":"3","year":"2020","journal-title":"NPJ digital medicine"},{"issue":"3","key":"2023040612500572000_2023.04.03.23288094v1.23","doi-asserted-by":"crossref","first-page":"317","DOI":"10.1016\/j.bandl.2013.10.002","article-title":"Speech in spinocerebellar ataxia","volume":"127","year":"2013","journal-title":"Brain and language"},{"key":"2023040612500572000_2023.04.03.23288094v1.24","doi-asserted-by":"publisher","DOI":"10.1080\/14015430600789203"},{"issue":"1","key":"2023040612500572000_2023.04.03.23288094v1.25","doi-asserted-by":"crossref","first-page":"126","DOI":"10.1007\/s12311-019-01084-8","article-title":"Speech and language disorders in friedreich ataxia: highlights on phenomenology, assessment, and therapy","volume":"19","year":"2020","journal-title":"The Cerebellum"},{"issue":"2","key":"2023040612500572000_2023.04.03.23288094v1.26","doi-asserted-by":"crossref","first-page":"e194","DOI":"10.1212\/WNL.0000000000009776","article-title":"Features of speech and swallowing dysfunction in pre-ataxic spinocerebellar ataxia type 2","volume":"95","year":"2020. [","journal-title":"Neurology"},{"key":"2023040612500572000_2023.04.03.23288094v1.27","doi-asserted-by":"publisher","DOI":"10.1044\/jslhr.4305.1275"},{"key":"2023040612500572000_2023.04.03.23288094v1.28","doi-asserted-by":"crossref","unstructured":"B. Kashyap , P. N. Pathirana , M. Horne , L. Power , and D. Szmulewicz , \u201cAutomated tongue-twister phrase-based screening for cerebellar ataxia using vocal tract biomarkers\u00a1sup\u00bf*\u00a1\/sup\u00bf,\u201d in 2019 41st Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC), 2019, pp. 7173\u20137176.","DOI":"10.1109\/EMBC.2019.8857868"},{"key":"2023040612500572000_2023.04.03.23288094v1.29","unstructured":"T. Sainburg , \u201ctimsainb\/noisereduce: v1.0,\u201d Jun. 2019. [Online]. Available: https:\/\/doi.org\/10.5281\/zenodo.3243139"},{"key":"2023040612500572000_2023.04.03.23288094v1.30","doi-asserted-by":"crossref","unstructured":"T. Sainburg , M. Thielk , and T. Q. Gentner , \u201cFinding, visualizing, and quantifying latent structure across diverse animal vocal repertoires,\u201d PLoS computational biology, vol. 16, no. 10, p. e1008228, 2020.","DOI":"10.1371\/journal.pcbi.1008228"},{"key":"2023040612500572000_2023.04.03.23288094v1.31","first-page":"18","article-title":"librosa: Audio and music signal analysis in python","volume":"8","year":"2015","journal-title":"in Proceedings of the 14th python in science conference"},{"key":"2023040612500572000_2023.04.03.23288094v1.32","unstructured":"K. He , X. Zhang , S. Ren , and J. Sun , \u201cDeep residual learning for image recognition,\u201d CoRR, vol. abs\/1512.03385, 2015. [Online]. Available: http:\/\/arxiv.org\/abs\/1512.03385"},{"key":"2023040612500572000_2023.04.03.23288094v1.33","article-title":"Adam: A Method for Stochastic Optimization","year":"2014","journal-title":"arXiv e-prints"},{"key":"2023040612500572000_2023.04.03.23288094v1.34","doi-asserted-by":"crossref","unstructured":"H. Zhou , H. Nguyen , A. Enriquez , L. Morsy , M. Curtis , T. Piser , C. Kenney , C. D. Stephen , A. S. Gupta , J. D. Schmahmann et al., \u201cAssessment of gait and balance impairment in people with spinocerebellar ataxia using wearable sensors,\u201d Neurological Sciences, pp. 1\u201311, 2021.","DOI":"10.1007\/s10072-021-05657-6"},{"key":"2023040612500572000_2023.04.03.23288094v1.35","unstructured":"H. Zhang , M. Cisse , Y. N. Dauphin , and D. Lopez-Paz , \u201cmixup: Beyond empirical risk minimization,\u201d 2018."},{"key":"2023040612500572000_2023.04.03.23288094v1.36","article-title":"Axiomatic Attribution for Deep Networks","year":"2017","journal-title":"arXiv e-prints"}],"container-title":[],"original-title":[],"link":[{"URL":"https:\/\/syndication.highwire.org\/content\/doi\/10.1101\/2023.04.03.23288094","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,1,15]],"date-time":"2026-01-15T19:31:09Z","timestamp":1768505469000},"score":1,"resource":{"primary":{"URL":"http:\/\/medrxiv.org\/lookup\/doi\/10.1101\/2023.04.03.23288094"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2023,4,4]]},"references-count":36,"URL":"https:\/\/doi.org\/10.1101\/2023.04.03.23288094","relation":{"is-preprint-of":[{"id-type":"doi","id":"10.1109\/ACCESS.2024.3393243","asserted-by":"subject"}]},"subject":[],"published":{"date-parts":[[2023,4,4]]},"subtype":"preprint"}}