Parkinson’s disease diagnostics using AI and natural language knowledge transfer

• Autorzy: Chronowski Maurycy , Kłaczyński Maciej , Dec-Ćwiek Małgorzata , Porębska Karolina

Identyfikatory

DOI

10.29354/diag/176931

• Języki publikacji: EN

Abstrakty

EN

With global life expectancy rising every year, ageing-associated diseases are becoming an increasingly important problem. Very often, successful treatment relies on early diagnosis. In this work, the issue of Parkinson's disease (PD) diagnostics is tackled. It is particularly important, as there are no certain antemortem methods of diagnosing PD - meaning that the presence of the disease can only be confirmed after the patient's death. In our work, we propose a non-invasive approach for classification of raw speech recordings for PD recognition using deep learning models. The core of the method is an audio classifier using knowledge transfer from a pretrained natural language model, namely wav2vec 2.0. The model was tested on a group of 38 PD patients and 10 healthy persons above the age of 50. A dataset of speech recordings acquired using a smartphone recorder was constructed and the recordings were labelled as PD/non-PD with the severity of the disease additionally rated using Hoehn-Yahr scale. We then benchmarked the classification performance against baseline methods. Additionally, we show an assessment of human-level performance with neurology professionals.

Słowa kluczowe

EN

Parkinson’s disease; digital diagnostics; artificial intelligence; speech processing

• Wydawca: Polskie Towarzystwo Diagnostyki Technicznej
• Czasopismo: Diagnostyka
• Rocznik: 2024
• Tom: Vol. 25, No. 1
• Strony: art. no. 2024103
• Opis fizyczny: Bibliogr. 15 poz., rys., tab.

Twórcy

autor

Chronowski Maurycy

• AGH University of Science and Technology, Adama Mickiewicza 30, 30-059 Kraków, Poland

• https://orcid.org/0009-0000-5715-8317

autor

Kłaczyński Maciej

• AGH University of Science and Technology, Adama Mickiewicza 30, 30-059 Kraków, Poland

• https://orcid.org/0000-0002-2009-1216

autor

Dec-Ćwiek Małgorzata

• Jagiellonian University, Collegium Medicum, Jakubowskiego 2, 30-688, Kraków, Poland

• https://orcid.org/0000-0002-9829-3301

autor

Porębska Karolina

• Jagiellonian University, Collegium Medicum, Jakubowskiego 2, 30-688, Kraków, Poland

• https://orcid.org/0000-0003-2494-4703

Bibliografia

• 1. Almeida JS, Rebouças Filho PP, Carneiro T, Wei W. Detecting Parkinson’s disease with sustained phonation and speech signals using machine learning techniques. Pattern Recognition Letters 2019; 125: 55-62. https://doi.org/10.1016/j.patrec.2019.04.005.
• 2. Babenko B. Multiple instance learning: Algorithms and applications. 01 2008.
• 3. Boualoulou N, Mounia M, Nsiri B, Behoussine Drissi T. A novel Parkinson's disease detection algorithm combined EMD, BFCC, and SVM classifier. Diagnostyka. 2023; 24(4): 2023404. https://doi.org/10.29354/diag/171712.
• 4. Chronowski M, Kłaczyński M, Dec-Ćwiek M. Speech and tremor tester - monitoring of neurodegenerative diseases using smartphone technology. Diagnostyka 2020;21(2):31-39. https://doi.org/10.29354/diag/122335.
• 5. Han Z, Tian R, Ren P, Zhou W, Wang P, Luo M. Parkinson’s disease and Alzheimer’s disease: a Mendelian randomization study. BMC Med Genet 2018;19. https://doi.org/10.1186/s12881-018-0721-7.
• 6. Jordal I, Nishi K, Bredin H. asteroidteam/torchaudiomentations:v0.10.1, 2022. https://doi.org/10.5281/zenodo.6381721.
• 7. Kłaczyński M. Vibroacoustic methods in diagnosis of selected laryngeal diseases. Journal of Vibroengineering 2015; 17(4): 2089-2098.
• 8. Liaqat A, Ce Z, Mingyi Z. Early diagnosis of Parkinson’s disease from multiple voice recordings by simultaneous sample and feature selection. Expert Systems with Applications 2019; 137: 22-28. https://doi.org/10.1016/j.eswa.2019.06.052.
• 9. Mąka J. The Polish linguistic test review in the assessment of Central Auditory Processing Disorders. Investigationes Linguisticae 2009; 18: 55. https://doi.org/10.14746/il.2009.18.4.
• 10. Reich SG. Does this patient have parkinson disease or essential tremor? Clinics in Geriatric Medicine 2020; 36(1): 25-34. https://doi.org/10.1186/s12881-018-0721-7.
• 11. Ribeiro MT, Singh S, Guestrin C. "Why should trust you?": Explaining the predictions of any classifier. 2016. https://doi.org/10.48550/ARXIV.1602.04938.
• 12. Sakar B, Isenkul M, Sakar CO. Collection and analysis of a Parkinson speech dataset with multiple types of sound recordings. Biomedical and Health Informatics, IEEE Journal of 2013; 17: 828-834. https://doi.org/10.1109/JBHI.2013.2245674.
• 13. Schneider S, Baevski A, Collobert R, Auli M. wav2vec: Unsupervised pre-training for speech recognition. arXiv:1904.05862 [cs], September 2019. https://doi.org/10.48550/arXiv.1904.05862.
• 14. Signaevsky M, Marami B, Prastawa M. Antemortem detection of Parkinson’s disease pathology in peripheral biopsies using artificial intelligence. acta neuropathol commun 2022; 10(21). https://doi.org/10.1186/s40478-022-01318-7.
• 15. Sveinbjornsdottir S, Marami B, Prastawa M. The clinical symptoms of Parkinson’s disease. Journal of Neurochemistry 2016; 139(S1): 318-324. https://doi.org/10.1111/jnc.13691. 1.

Uwagi

Opracowanie rekordu ze środków MEiN, umowa nr SONP/SP/546092/2022 w ramach programu "Społeczna odpowiedzialność nauki" - moduł: Popularyzacja nauki i promocja sportu (2022-2023).