Advanced signal processing methods for analysis of respiratory sounds

• Main Author: Jin, Feng
• Other Authors: Farook Sattar
• Format: Theses and Dissertations
• Language: English
• Published: 2010
• Subjects: DRNTU::Engineering::Electrical and electronic engineering::Control and instrumentation::Medical electronics
• Online Access: https://hdl.handle.net/10356/36292
• Institution: Nanyang Technological University

Respiratory sounds (RSs) auscultation using stethoscope is often the first noninvasive resource available to physicians for the detection and analysis of respiratory ailments. Motivated by the subjectivity of human auditory system, computerized auscultation combined with signal processing techniques have boosted the diagnostic capabilities of RS. This PhD thesis concentrates on the advanced signals processing methods for RS analysis from several aspects including: RS denoising by means of heart sounds (HSs) cancellation (Part I), respiratory phase detection (Part II), as well as RS analysis and feature extraction methods for RS classification (Part III). Part I solves the problem of HS cancellation through a two-steps scheme consisting of HS localization and HS removal from different types of single-channel RS recordings.