Related Experiment Videos
Time-scale segmentation of respiratory sounds
E Ademovic1, J C Pesquet, G Charbonneau
1Laboratoire des Signaux et Systèmes, CNRS/Univ. Paris-Sud and GdR-PRC ISIS, ESE, Gif sur Yvette, France. ademovic@lss.supelec.fr
Summary
Analyzing respiratory sounds is complex due to diverse signal characteristics. New wavelet packet decomposition methods offer improved time segmentation for better analysis of normal and adventitious respiratory sounds.
Area of Science:
- Pulmonology
- Biomedical Engineering
- Signal Processing
Background:
- Respiratory sounds encompass normal and adventitious events with varied characteristics.
- Analyzing these complex signals is challenging with conventional single techniques.
- Time-frequency and time-scale methods enhance signal analysis accuracy.
Purpose of the Study:
- To introduce novel approaches for respiratory sound segmentation.
- To leverage wavelet packet decomposition for improved signal analysis.
- To address the limitations of existing techniques in analyzing complex respiratory sounds.
Main Methods:
- Utilized wavelet packet decomposition (WPD) for signal analysis.
- Developed new WPD-based approaches for time segmentation.
- Focused on segmenting both normal and adventitious respiratory sounds.
Main Results:
- Demonstrated the efficacy of WPD for respiratory sound segmentation.
- Achieved accurate time segmentation of complex respiratory signals.
- New approaches provide enhanced analysis capabilities.
Conclusions:
- Wavelet packet decomposition offers a powerful tool for respiratory sound analysis.
- The presented methods improve the segmentation of respiratory sounds.
- These advancements facilitate a more comprehensive study of respiratory phenomena.