Szczegóły publikacji
Opis bibliograficzny
Parkinson's disease classification based on vowel sound / D. HEMMERLING, D. Sztahó // W: Models and analysis of vocal emissions for biomedical applications : 11th international workshop : December 17–19, 2019, Firenze, Italy / ed. by Claudia Manfredi. — Firenze : Firenze University Press, 2019. — (Proceedings e Report ; ISSN 2704-601X ; 122). — ISBN: 978-88-6453-951-5; e-ISBN: 978-88-6453-961-4. — S. 29–32. — Bibliogr. s. 32, Abstr.
Autorzy (2)
- AGHHemmerling Daria
- Sztahó D.
Słowa kluczowe
Dane bibliometryczne
| ID BaDAP | 127438 |
|---|---|
| Data dodania do BaDAP | 2020-02-19 |
| Rok publikacji | 2019 |
| Typ publikacji | materiały konferencyjne (aut.) |
| Otwarty dostęp |  |
| Creative Commons | |
| Wydawca | Università degli Studi di Firenze |
| Czasopismo/seria | Proceedings e Report |
Abstract
Parkinson’s disease (PD) is the second most frequent neurodegenerative disorder that causes decisive deterioration of the quality of life through severe motor and cognitive dysfunctions. In this study we present the usage of vowel signals to accurately detect PD by using machine learning methods. The data set consists in total of 198 recordings of vowel /a/ phonated in sustained manner, where 50% of data was assigned as a representation of Parkinson's disease state. The voice signals are described by the set of features extracted from time, frequency and cepstral domains applied to Principal Component Analysis (PCA) and nonlinear Support Vector Machine (SVM) to distinguish between PD patients and healthy control group. The results ensure 93.43% of classification accuracy. © 2019 FUP,
