Related Experiment Videos
Microphone and electroglottographic data from dysphonic patients: type 1, 2 and 3 signals
A Behrman1, C J Agresti, E Blumstein
1Department of Otolaryngology, Long Island Jewish Medical Center, New York, New York, USA.
Summary
Jitter and shimmer analysis may be unreliable for dysphonic voice signals. This study found many non-periodic signals, suggesting voice signal typing is crucial for accurate analysis.
Area of Science:
- Speech science
- Acoustic analysis
- Biomedical engineering
Background:
- Traditional voice analysis metrics like jitter and shimmer assume signal periodicity.
- This assumption may limit their validity for complex vocal fold dynamics in dysphonia.
Purpose of the Study:
- To investigate the prevalence of different signal types (periodic, bifurcating, chaotic) in dysphonic voices.
- To assess the necessity of signal classification prior to acoustic analysis.
Main Methods:
- Analysis of microphone and electroglottographic signals from 202 patients with dysphonia.
- Classification of signals into types based on periodicity, bifurcations, and chaotic behavior.
Main Results:
- 42% of signals were nearly periodic (Type 1).
- 35% exhibited bifurcations or modulations (Type 2).
- 22% were chaotic (Type 3), with difficulty distinguishing Type 2 from Type 3 in 40% of cases.
Conclusions:
- A significant proportion of dysphonic voice signals are non-periodic.
- Signal typing is an essential preliminary step for accurate analysis of voice data.
- Current acoustic analysis methods may require adaptation for non-periodic signals.