Feature extraction of photoplethysmography signal using wavelet approach

Photoplethysmography (PPG) signal has been commonly used in detecting the peripheral blood volume pulse. This kind of optical-based technique is simple, rapid, low-cost and non-invasive. It can be used widely for different purposes. Many useful information can be obtained from the PPG signal, including heart rate, respiratory rate, pulse wave velocity (PWV) and so on. In order to obtain these information, feature points must be extracted first. In this paper eight healthy subjects were studied. One PPG probe was used to record the PPG signal from left ear and the ECG signal has also been recorded simultaneously. These signals were converted into digital signals and stored into computer for off-line analysis. An automatic algorithm based on wavelet approach has been developed to extract feature points from the PPG signal, then compare with the manual measurement. The result shows that the automatic algorithm provided an effective way for PPG signal feature extraction.