Development for Evaluation of Pathological Voices (rough, voice tremor) aimed to support of Clinical Inspection.
Grant-in-Aid for General Scientific Research (C)
|Allocation Type||Single-year Grants|
Biomedical engineering/Biological material science
|Research Institution||Kanazawa Institute of Technology|
HIRAMA Junji Kanazawa Institute of Technology, Department of Engeneering, Lecturer, 工学部, 講師 (40181185)
|Project Period (FY)
1994 – 1995
Completed(Fiscal Year 1995)
|Budget Amount *help
¥2,200,000 (Direct Cost : ¥2,200,000)
Fiscal Year 1995 : ¥1,400,000 (Direct Cost : ¥1,400,000)
Fiscal Year 1994 : ¥800,000 (Direct Cost : ¥800,000)
|Keywords||pathological voice / rough / voice tremor / fluctuation of fundamental frequency / synthetic voice / acoustical impression / clinical inspection / evaluation system|
In this paper. Author investigated effect of the fluctuation of fundamental frequency (F _0) on the identification of pathological of two kinds of voice qualities, rough and voice tremor.
The aim of this study is to examine the relationships between the acoustic characteristics and psychoacoustic judgment on the basis of the fluctuation of F _0.
And the acoustic impression training system and voice diagnosis equipment (clinical examination) of pathological voices for the "laryngologists" were developed.
The procedures and finding in this study are summerized as follows.
(1) Characteristics of spectral peakes of characteristics shapes in spectrum domein of fluctuations of the time series, for fundamental frequency (F _0), were examined for a pathological voice which was psychoacoustically evaluated as "mainly rough and voice tremor".
(2) Synthetic vowels/a/with frequency modulated F _0 fluctuation (FM model) were generated for systematic combination of modulation parameters.
And the time series of fluctuations of F _0 can be generated from the "Slope-Peak model" as the first-order approximation.
Also, by using the each model, synthetic vowel/a/with F _0 fluctuation were generated respectively.
Subjects were asked to listen to synthetic voices and evaluated pyschoacoustic categories (rough and voice tremor).
The following results were obtained.
(1) Relationships between the parameters for each model ( "FM" and "Slope-Peak" model) and the categories for the pathological voices were found.
(2) The acoustic impression training system and voice clinical examination system of pathological voice for "laryngologists" would be attractive in actual clinical inspection.
These results would be usefull for medical electronics and biological engineering, also to clinical examination of voice.
Research Output (10results)