Output details
11 - Computer Science and Informatics
University of Surrey
Localizing Heart Sounds in Respiratory Signals Using Singular Spectrum Analysis
<12>To diagnose lung abnormalities, heart sound has to be cancelled from combined heart and lung sounds recorded by electronic stethoscopes. The novelty is that singular spectrum analysis, previously used in econometrics, has been implemented to enable, for the first time, identification of heart sound segments and their temporal locations within mixed signals. This approach is beneficial for clinicians and patients in clinical diagnosis of various lung diseases. It has also opened a new research direction to automatically detect these abnormalities. The work has been extended to exploit cyclostationarity of the heart sound in more accurate separation from the lung sound.