Analysis of acoustic characteristics of speech in Parkinson's disease patient recorded via smartphones
10.3969/j.issn.1006-7299.2025.06.010
- VernacularTitle:智能手机采录帕金森病患者语音的声学特征分析
- Author:
Jiaxing ZHENG
1
;
Menglan HAN
;
Gang WANG
;
Yuting TANG
;
Chengju LIAO
;
Kaiwen CHEN
;
Weixin LING
;
Zhuoming CHEN
Author Information
1. 暨南大学附属第一医院康复医学科(广州 510630)
- Publication Type:Journal Article
- Keywords:
Parkinson's disease;
Formants;
Acoustic analysis;
Fundamental frequency;
Smartphone
- From:
Journal of Audiology and Speech Pathology
2025;33(6):560-563
- CountryChina
- Language:Chinese
-
Abstract:
Objective To investigate the acoustic characteristics of speech in Parkinson's disease(PD)pa-tients compared to healthy individuals using smartphone-recorded speech,laying the foundation for future intelligent cloud-based PD detection.Methods A total of 81 PD patients and 63 healthy controls were randomly selected to re-cord their speech using Huawei,Xiao mi,and Apple smartphones through the"Active Health"screening app from the National Key R&D Program.Using the Praat library in Python,parameters such as mean fundamental frequen-cy(Mean F0),minimum fundamental frequency(Min F0),maximum fundamental frequency(Max F0),formants(F1,F2,F3,F4),shimmer,noise-to-harmonics ratio(NHR),and jitter were extracted and compared.Results In PD patients,jitter and shimmer were significantly higher in the/ba/,/bi/,/bu/,and/bo/vowels compared to the control group,while Mean F0 and Max F0 was significantly lower in/bi/and/bo/.Min F0 was significantly higher in/bi/and/bu/for PD patients.No significant differences were found in F1,F2,F3,or F4 between PD pa-tients and controls in/ba/,/bi/,/bo/and/bu/.Conclusion Quantitative analysis of smartphone-recorded speech can detect speech changes in PD patients,providing a solid foundation for the development of a cloud-based PD de-tection system.
