Related Experiment Videos
Speech audiometric trial using synthetic vowels produced with DEC-talk
1Department of Otolaryngology, University of Tokushima, School of Medicine, Japan.
Medical Informatics = Medecine Et Informatique
|October 1, 1990
Summary
Synthesized speech sounds, including distorted ones, achieved high articulation scores in speech audiometry. This novel method offers a reliable alternative for assessing hearing loss during otological examinations.
Area of Science:
- Audiology
- Speech Science
- Biomedical Engineering
Background:
- Speech audiometry is crucial for diagnosing hearing loss.
- Current methods using natural speech recordings face challenges with stimulus consistency and availability.
- Developing standardized acoustic stimuli is essential for accurate audiological assessments.
Purpose of the Study:
- To evaluate the efficacy of synthesized speech sounds for speech audiometric testing.
- To assess the degree of hearing loss and its pathogenesis using novel acoustic stimuli.
- To explore a potential solution for limitations in using recorded natural speech.
Main Methods:
- Speech sounds, including vowels and distorted speech, were synthesized using DECtalk.
- These synthesized stimuli were employed in speech audiometric trials.
- Articulation scores were measured to evaluate the effectiveness of the synthesized sounds.
Main Results:
- Undistorted synthesized vowels achieved 100% articulation.
- Distorted synthetic speech signals yielded reasonable articulation scores.
- The synthesized speech method demonstrated potential for audiological applications.
Conclusions:
- Synthesized speech sounds are effective acoustic stimuli for speech audiometry.
- This method offers a viable alternative to recorded natural speech, addressing existing limitations.
- The approach holds promise for improving the accuracy and consistency of hearing loss evaluations.