Related Experiment Videos
Inverse filtering applied to upper airway sounds
1Department of Electrical Engineering and Electronics, Liverpool University, UK.
Summary
Inverse filtering analyzes airway sounds to detect changes in vocal tract dimensions and assess voice disorders. This digital signal processing technique aids in diagnosing conditions like sleep apnea and hoarseness.
Area of Science:
- Digital signal processing
- Speech science
- Medical diagnostics
Background:
- Upper airway cavities introduce resonances into speech sounds.
- Inverse filtering digitally removes these resonances to isolate the excitation source.
- Filter parameters from linear prediction analysis reveal resonance frequencies and bandwidths.
Purpose of the Study:
- To outline the principle of inverse filtering.
- To describe two diagnostic applications of inverse filtering for upper airway sounds.
- To investigate non-invasive measurement of airway dimension changes and vocal tract abnormalities.
Main Methods:
- Application of inverse filtering to speech-like sounds.
- Linear prediction analysis to compute filter parameters.
- Analysis of residual signals for diagnostic insights.
Main Results:
- Inverse filtering can measure resonance frequency shifts (approx. 10% in normals) due to posture changes.
- These measurements are valuable for assessing sleep apnea.
- Residual signal parameters show potential for evaluating laryngeal infection and hoarseness severity.
Conclusions:
- Inverse filtering is a viable technique for non-invasive upper airway sound analysis.
- It offers diagnostic potential for conditions affecting airway dimensions and vocal fold function.
- Further research supports its use in sleep apnea assessment and hoarseness evaluation.