How Neurological Disorders Affect Acoustic Features of Speech?
A Data Mining Approach
Keywords:
Neurological Disorders, Acoustic Features, Speech, Data MiningAbstract
Introduction: Neurological disorders often manifest as alterations in speech production, affecting acoustic features such as pitch, rhythm, articulation, and voice quality. These speech changes can serve as non-invasive biomarkers for early diagnosis and monitoring. This study aims to investigate how various neurological disorders affect the acoustic features of speech across multiple speaking styles using a data mining approach.
Material and Methods: We collected speech recordings from 383 participants, including 264 individuals diagnosed with different neurological disorders and 119 healthy controls. Participants read a standardized text in four speaking styles: questioning, excited, angry, and happy. A comprehensive set of acoustic features, covering prosodic, spectral, voice quality, formant, and temporal parameters, was extracted. Deep learning models were trained separately for each speaking style to classify neurological status. Feature importance analyses were conducted to identify key acoustic indicators of neurological impairment.
Results: The deep learning models achieved classification accuracies ranging from 82.7% to 87.9% across speaking styles, with the highest performance observed in the angry speech condition. Prosodic features, particularly fundamental frequency and speaking rate, alongside voice quality measures such as jitter and shimmer, emerged as the most discriminative features. Distinct acoustic profiles were identified for different neurological disorders, and several features correlated significantly with clinical severity scores. Speaking style influenced the detectability of speech impairments, underscoring the value of analyzing diverse speech contexts.
Conclusion: Our findings demonstrate that neurological disorders induce characteristic alterations in acoustic speech features that can be effectively captured using data mining and deep learning techniques. Incorporating multiple speaking styles enhances diagnostic sensitivity. This approach holds promise for developing accessible, non-invasive speech-based biomarkers to support early diagnosis and monitoring of neurological diseases.
References
1. Feigin VL, Nichols E, Alam T, Bannick MS, Beghi E, Blake N, et al. Global, regional, and national burden of neurological disorders, 1990–2016: A systematic analysis for the global burden of disease study 2016. Lancet Neurol. 2019; 18(5): 459-80. PMID: 30879893 DOI: 10.1016/S1474-4422(18)30499-X
2. Feigin VL, Vos T, Nichols E, Owolabi MO, Carroll WM, Dichgans M, et al. The global burden of neurological disorders: Translating evidence into policy. Lancet Neurol. 2020; 19(3): 255-65. PMID: 31813850 DOI: 10.1016/S1474-4422(19)30411-9
3. Hecker P, Steckhan N, Eyben F, Schuller BW, Arnrich B. Voice analysis for neurological disorder recognition–a systematic review and perspective on emerging trends. Front Digit Health. 2022; 4: 842301. PMID: 35899034 DOI: 10.3389/fdgth.2022.842301
4. Mayer G, Happe S, Evers S, Hermann W, Jansen S, Kallweit U, et al. Insomnia in neurological diseases. Neurol Res Pract. 2021; 3: 1-12. PMID: 33691803 DOI: 10.1186/s42466-021-00106-3
5. Gilmour GS, Nielsen G, Teodoro T, Yogarajah M, Coebergh JA, Dilley MD, et al. Management of functional neurological disorder. J Neurol. 2020; 267: 2164-72. PMID: 32193596 DOI: 10.1007/s00415-020-09772-w
6. Ding C, Wu Y, Chen X, Chen Y, Wu Z, Lin Z, et al. Global, regional, and national burden and attributable risk factors of neurological disorders: The global burden of disease study 1990–2019. Front Public Health. 2022; 10: 952161. PMID: 36523572 DOI: 10.3389/fpubh.2022.952161
7. Zhang R, Liu H, Pu L, Zhao T, Zhang S, Han K, et al. Global burden of ischemic stroke in young adults in 204 countries and territories. Neurology. 2023; 100(4): e422-34. PMID: 36307222 DOI: 10.1212/WNL.0000000000201467
8. Steinmetz JD, Seeher KM, Schiess N, Nichols E, Cao B, Servili C, et al. Global, regional, and national burden of disorders affecting the nervous system, 1990–2021: A systematic analysis for the global burden of disease study 2021. Lancet Neurol. 2024; 23(4): 344-81. PMID: 38493795 DOI: 10.1016/S1474-4422(24)00038-3
9. Saeedi S, Hetjens S, Grimm M, Latoszek BBv. Acoustic speech analysis in Alzheimer’s disease: A systematic review and meta-analysis. J Prev Alzheimers Dis. 2024; 11(6): 1789-97. PMID: 39559890 DOI: 10.14283/jpad.2024.132
10. Cho S, Cousins KAQ, Shellikeri S, Ash S, Irwin DJ, Liberman MY, et al. Lexical and acoustic speech features relating to Alzheimer disease pathology. Neurology. 2022; 99(4): e313-22. PMID: 35487701 DOI: 10.1212/WNL.0000000000200581
11. Noffs G, Boonstra FM, Perera T, Kolbe SC, Stankovich J, Butzkueven H, et al. Acoustic speech analytics are predictive of cerebellar dysfunction in multiple sclerosis. Cerebellum. 2020; 19: 691-700. PMID: 32556973 DOI: 10.1007/s12311-020-01151-5
12. He D, Feenaughty L, Wan Q. Global acoustic speech temporal characteristics for Mandarin speakers with Parkinson’s disease during syllable repetition and passage reading. Am J Speech Lang Pathol. 2023; 32(5): 2232-44. PMID: 37625136 DOI: 10.1044/2023_AJSLP-23-00062
13. Krýže P, Tykalová T, Růžička E, Rusz J. Effect of reading passage length on quantitative acoustic speech assessment in Czech-speaking individuals with Parkinson’s disease treated with subthalamic nucleus deep brain stimulation. J Acoust Soc Am. 2021; 149(5): 3366-74. PMID: 34241103 DOI: 10.1121/10.0005050
14. Kim KS, Wang H, Max L. It's about time: Minimizing hardware and software latencies in speech research with real-time auditory feedback. J Speech Lang Hear Res. 2020; 63(8): 2522-34. PMID: 32640180 DOI: 10.1044/2020_JSLHR-19-00419
15. Rusz J, Krack P, Tripoliti E. From prodromal stages to clinical trials: The promise of digital speech biomarkers in Parkinson’s disease. Neurosci Biobehav Rev. 2024; 167: 105922. PMID: 39424108 DOI: 10.1016/j.neubiorev.2024.105922
16. Dorsey ER, Omberg L, Waddell E, Adams JL, Adams R, Ali MR, et al. Deep phenotyping of Parkinson’s disease. J Parkinsons Dis. 2020; 10(3): 855-73. PMID: 32444562 DOI: 10.3233/JPD-202006
17. Livezey JA, Glaser JI. Deep learning approaches for neural decoding across architectures and recording modalities. Brief Bioinform. 2021; 22(2): 1577-91. PMID: 33372958 DOI: 10.1093/bib/bbaa355
18. Thies T, Mallick E, Tröger J, Baykara E, Mücke D, Barbe MT. Automatic speech analysis combined with machine learning reliably predicts the motor state in people with Parkinson’s disease. NPJ Parkinsons Dis. 2025; 11(1): 105. PMID: 40316531 DOI: 10.1038/s41531-025-00959-4
19. Huang Y-J, Lin Y-T, Liu C-C, Lee L-E, Hung S-H, Lo J-K, et al. Assessing schizophrenia patients through linguistic and acoustic features using deep learning techniques. IEEE Trans Neural Syst Rehabil Eng. 2022; 30: 947-56. PMID: 35358049 DOI: 10.1109/TNSRE.2022.3163777
20. Sayadi M, Varadarajan V, Langarizadeh M, Bayazian G, Torabinezhad F. A systematic review on machine learning techniques for early detection of mental, neurological and laryngeal disorders using patient’s speech. Electronics. 2022; 11(24): 4235.
21. Sayadi MJ, Langarizadeh M, Torabinezhad F, Bayazian G. Voice as an indicator for laryngeal disorders using data mining approach. Frontiers in Health Informatics. 2024; 13: 205.
Downloads
Published
Issue
Section
License
Copyright (c) 2025 Advances in Medical Informatics
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.
