Hodnocení poruch artikulace pacientů s Parkinsonovou nemocí založené na rezonancích v hlasovém traktu
Loading...
Date
2022-03-23
Authors
Kováč, Daniel
ORCID
Advisor
Referee
Mark
Journal Title
Journal ISSN
Volume Title
Publisher
International Society for Science and Engineering, o.s.
Abstract
Parkinsonova nemoc (PN) je často asociována s hypokinetickou dysartrií, což je porucha motorické realizace řeči, která se mj. manifestuje v artikulaci. Vlivem rigidity artikulačních orgánů tak např. dochází ke snížení srozumitelnosti řečového projevu. Cílem této studie je návrh nových řečových parametrů, které by poruchy artikulace dokázaly kvantifikovat a také by umožnily diskriminovat zdravé řečníky od pacientů s PN. Byl ověřen útlum na rezonanční frekvenci (RFA) a dálší dva akustické parametry řeči založené na míře rezonancí v nadhrtanových dutinách byly navrženy. Jejich diskriminační síla byla testována na databázi čítající 19 zdravých kontrol (ZK) a 28 pacientů s PN. Na základě Studentova t-testu bylo možné pozorovat signifikantní rozdíly mezi ZK a PN u všech tří parametrů. Pomocí logistické regrese (leave- one-out křížová validace) bylo možné PN klasifikovat s 89% senzitivitou a 96% specificitou. Tato pilotní studie naznačuje, že rezonance v nadhrtanových dutinách mohou hrát významnou roli při podpůrné diagnóze PN z řeči.
Parkinson's disease (PD) is often associated with hypokinetic dysarthria, a disorder of motor realization of speech that manifests, among other areas, in articulation. For example, the rigidity of the articulatory organs results in a reduction in speech intelligibility. This study aims to propose new speech features that can quantify articulation disorders and allow the discrimination of healthy speakers from PD patients. Resonant frequency attenuation (RFA) was validated, and two additional acoustic speech features based on the degree of resonances in the supraglottic cavities were designed. Their discriminative power was tested on a database comprising 19 healthy controls (HC) and 28 PD patients. Based on Student's t-test, significant differences between HC and PD could be observed for all three features. Using logistic regression (leave-one-out cross-validation), PD could be classified with 89% sensitivity and 96% specificity. This pilot study suggests that supraglottic resonances may play a significant role in supporting the diagnosis of PD from speech.
Parkinson's disease (PD) is often associated with hypokinetic dysarthria, a disorder of motor realization of speech that manifests, among other areas, in articulation. For example, the rigidity of the articulatory organs results in a reduction in speech intelligibility. This study aims to propose new speech features that can quantify articulation disorders and allow the discrimination of healthy speakers from PD patients. Resonant frequency attenuation (RFA) was validated, and two additional acoustic speech features based on the degree of resonances in the supraglottic cavities were designed. Their discriminative power was tested on a database comprising 19 healthy controls (HC) and 28 PD patients. Based on Student's t-test, significant differences between HC and PD could be observed for all three features. Using logistic regression (leave-one-out cross-validation), PD could be classified with 89% sensitivity and 96% specificity. This pilot study suggests that supraglottic resonances may play a significant role in supporting the diagnosis of PD from speech.
Description
Keywords
Citation
Elektrorevue. 2022, vol. 24, č. 1, s. 29-34. ISSN 1213-1539
http://elrevue.utko.feec.vutbr.cz/index.php/Elektrorevue/article/view/182/190
http://elrevue.utko.feec.vutbr.cz/index.php/Elektrorevue/article/view/182/190
Document type
Peer-reviewed
Document version
Published version
Date of access to the full text
Language of document
cs
Study field
Comittee
Date of acceptance
Defence
Result of defence
Document licence
(C) 2022 Elektrorevue