Related Experiment Videos
Laryngeal perturbation analysis: minimum length of analysis window
1Center for Speech and Swallowing Disorders, Pritzker School of Medicine, University of Chicago.
Journal of Speech and Hearing Research
|June 1, 1991
Summary
Determining the optimal signal length for laryngeal perturbation analysis is crucial. This study found that voice analysis requires more cycles for jitter and shimmer to stabilize, especially for pathological voices.
Area of Science:
- Acoustic analysis
- Voice science
- Speech pathology
Background:
- Laryngeal perturbation measures analyze voice signal cycle-to-cycle changes.
- Widespread clinical use contrasts with methodological disagreements, particularly signal length.
- Previous research lacks consensus on optimal analysis window for perturbation measures.
Purpose of the Study:
- To investigate how laryngeal perturbation measures change with varying signal lengths.
- To determine the necessary number of cycles for stable jitter and shimmer measurements.
- To compare analysis window requirements for pathological versus nonpathological voices.
Main Methods:
- Analysis of acoustic voice signals from 18 subjects with suspected laryngeal dysfunction.
- Systematic examination of laryngeal perturbation measures (jitter, shimmer) across different signal durations.
- Calculation of asymptote points for jitter and shimmer based on cycle count.
Main Results:
- Jitter measures required up to 190 cycles to reach asymptote.
- Shimmer measures required up to 130 cycles to reach asymptote.
- Pathological voices indicated a need for longer analysis windows compared to nonpathological voices.
Conclusions:
- Sufficient signal length is critical for reliable laryngeal perturbation analysis.
- Current methodologies may underestimate stability due to insufficient analysis windows.
- Clinical application of perturbation measures necessitates standardized, longer analysis durations for accuracy.