Speech-Based Artificial Intelligence for Early Parkinson's Disease: A Framework for Language, Recording, and Clinical Robustness

Abstract Voice recording offers an unusually accessible way to investigate motor function, but accessibility does not establish diagnostic specificity. A classifier can separate recordings from people with and without Parkinson's disease while learning differences in language, microphone, recruitment, or treatment. These alternatives become especially consequential when the intended use shifts from recognizing established disease to identifying risk before diagnosis. This selected narrative review examines foundational acoustic studies, multilingual investigations, telephone datasets, and recent work using ordinary smartphone calls. It distinguishes three targets that are often conflated: detecting a speech abnormality, recognizing clinically diagnosed Parkinson's disease, and forecasting a future clinical diagnosis. The methodological contribution is a proposed robustness matrix that crosses the intended prediction target with participant independence, recording conditions, language transfer, clinical alternatives, and follow-up. A companion sampling design pairs a short standardized speech task with repeat recordings and contextual metadata, while retaining recording failure as an outcome. The framework emphasizes explicit abstention, comparison with simple baselines, participant-level uncertainty, and prospective evaluation at the prevalence of the intended setting. It also separates acoustic features from conversational content to make privacy choices visible. Published evidence supports further research on speech biomarkers, including in populations with isolated rapid eye movement sleep behavior disorder. It does not establish that an untested online recording service can diagnose Parkinson's disease. For NeuralCipher.ai, the appropriate contribution is a reproducible research workflow whose claims remain tied to the population and evaluation actually studied. Article details Authors: NeuralCipherai; Kadir Tamrak; Salih Yaldız; Feride Yaldız; Ömer Ağyol; Yavuz Selim Silay; Hasan Randa Publisher: neluracipher.ai DOI: 10.5281/zenodo.22779003 Version: 1.0Language: English Project website: https://neuralcipher.ai References Arora, S., Baghai-Ravary, L., & Tsanas, A. (2019). Developing a large scale population screening tool for the assessment of Parkinson's disease using telephone-quality voice. The Journal of the Acoustical Society of America, 145(5), 2871–2884. https://doi.org/10.1121/1.5100272 Arora, S., & Tsanas, A. (2021). Assessing Parkinson’s disease at scale using telephone-recorded speech: Insights from the Parkinson’s Voice Initiative. Diagnostics, 11(10), Article 1892. https://doi.org/10.3390/diagnostics11101892 Ho, A. K., Bradshaw, J. L., & Iansek, R. (2008). For better or worse: The effect of levodopa on speech in Parkinson's disease. Movement Disorders, 23(4), 574–580. https://doi.org/10.1002/mds.21899 Illner, V., Novotný, M., Kouba, T., Tykalová, T., Šimek, M., Sovka, P., Švihlík, J., Růžička, E., Šonka, K., Dušek, P., & Rusz, J. (2024). Smartphone voice calls provide early biomarkers of parkinsonism in rapid eye movement sleep behavior disorder. Movement Disorders, 39(10), 1752–1762. https://doi.org/10.1002/mds.29921 Little, M. A., McSharry, P. E., Hunter, E. J., Spielman, J., & Ramig, L. O. (2009). Suitability of dysphonia measurements for telemonitoring of Parkinson's disease. IEEE Transactions on Biomedical Engineering, 56(4), 1015–1022. https://doi.org/10.1109/tbme.2008.2005954 Orozco-Arroyave, J. R., Hönig, F., Arias-Londoño, J. D., Vargas-Bonilla, J. F., Daqrouq, K., Skodda, S., Rusz, J., & Nöth, E. (2016). Automatic detection of Parkinson's disease in running speech spoken in three different languages. The Journal of the Acoustical Society of America, 139(1), 481–500. https://doi.org/10.1121/1.4939739 Ramig, L., Halpern, A., Spielman, J., Fox, C., & Freeman, K. (2018). Speech treatment in Parkinson's disease: Randomized controlled trial (RCT). Movement Disorders, 33(11), 1777–1791. https://doi.org/10.1002/mds.27460 Rusz, J., Cmejla, R., Ruzickova, H., & Ruzicka, E. (2011). Quantitative acoustic measurements for characterization of speech and voice disorders in early untreated Parkinson’s disease. The Journal of the Acoustical Society of America, 129(1), 350–367. https://doi.org/10.1121/1.3514381 Rusz, J., Hlavnička, J., Novotný, M., Tykalová, T., Pelletier, A., Montplaisir, J., Gagnon, J.-F., Dušek, P., Galbiati, A., Marelli, S., Timm, P. C., Teigen, L. N., Janzen, A., Habibi, M., Stefani, A., Holzknecht, E., Seppi, K., Evangelista, E., Rassu, A. L., . . . Šonka, K. (2021). Speech biomarkers in rapid eye movement sleep behavior disorder and Parkinson disease. Annals of Neurology, 90(1), 62–75. https://doi.org/10.1002/ana.26085 Rusz, J., Hlavnička, J., Tykalová, T., Bušková, J., Ulmanová, O., Růžička, E., & Šonka, K. (2016). Quantitative assessment of motor speech abnormalities in idiopathic rapid eye movement sleep behaviour disorder. Sleep Medicine, 19, 141–147. https://doi.org/10.1016/j.sleep.2015.07.030 Šimek, M., Tykalová, T., Novotný, M., Illner, V., Kouba, T., Šubert, M., Sovka, P., Švihlík, J., Růžička, E., Šonka, K., Dušek, P., & Rusz, J. (2026). Speech biomarkers from smartphone calls track progression in REM sleep behavior disorder and Parkinson's disease. Annals of Neurology, 99(4), 935–948. https://doi.org/10.1002/ana.78140 Skodda, S., Grönheit, W., Mancinelli, N., & Schlegel, U. (2013). Progression of voice and speech impairment in the course of Parkinson's disease: A longitudinal study. Parkinson's Disease, 2013, Article 389195. https://doi.org/10.1155/2013/389195 Tracey, B., Volfson, D., Glass, J., Haulcy, R. M., Kostrzebski, M., Adams, J., Kangarloo, T., Brodtmann, A., Dorsey, E. R., & Vogel, A. (2023). Towards interpretable speech biomarkers: Exploring MFCCs. Scientific Reports, 13(1), Article 22787. https://doi.org/10.1038/s41598-023-49352-2 Tsanas, A., Little, M. A., McSharry, P. E., & Ramig, L. O. (2010). Accurate telemonitoring of Parkinson's disease progression by noninvasive speech tests. IEEE Transactions on Biomedical Engineering, 57(4), 884–893. https://doi.org/10.1109/tbme.2009.2036000 Tsanas, A., Little, M. A., McSharry, P. E., Spielman, J., & Ramig, L. O. (2012). Novel speech signal processing algorithms for high-accuracy classification of Parkinson's disease. IEEE Transactions on Biomedical Engineering, 59(5), 1264–1271. https://doi.org/10.1109/tbme.2012.2183367