Remote Pulse Rate Variability Measurement with Adaptive Slope Sum Function

Compared to Electrocardiography (ECG) and contact Photoplethysmography (PPG), the signal obtained from remote Photoplethysmography (rPPG) is much more noisy due to the body movement, light variation, sensor noise, low scanning frequency of the cameras, etc. As a result, the measurement of Pulse Rate...

Full description

Saved in:
Bibliographic Details
Published in2023 11th International Conference on Bioinformatics and Computational Biology (ICBCB) pp. 76 - 81
Main Authors Yang, Kehang, Li, Peixi
Format Conference Proceeding
LanguageEnglish
Published IEEE 21.04.2023
Subjects
Continuous Wavelet Transform
Electrocardiography
Electroencephalography
Emotion recognition
Peak Detection
Photoplethysmography
Pulse measurements
Pulse Rate Variability
Remote Photoplethysmography
Size measurement
Slope Sum Function
Online AccessGet full text

Cover

More Information
Summary:Compared to Electrocardiography (ECG) and contact Photoplethysmography (PPG), the signal obtained from remote Photoplethysmography (rPPG) is much more noisy due to the body movement, light variation, sensor noise, low scanning frequency of the cameras, etc. As a result, the measurement of Pulse Rate Variability (PRV) based on rPPG is a more challenging task than measuring PRV and Heart Rate Variability (HRV) from contact PPG and ECG, because it requires precise peak detection on the noisy rPPG signal. To address this issue, some researchers used slope sum function (SSF) to enhance the rising trend of the signal and reduce the downward trend of it, so that the signal became more clear and easier for peak detection. This method is relatively effective, but it has some disadvantages. Firstly, the window size of the SSF is set manually. Secondly, there is a shift between the peaks on the new signal generated by the SSF and the original signal. Thirdly, the parameter is fixed for the entire dataset, which is not ideal for real-life scenario where there are sudden changes in signal. In this paper, we proposed a novel algorithm which adaptively and automatically set the window size of SSF using continuous wavelet transform. We experimentally showed that the proposed method improved the remote measurement of PRV compared to the existing methods in the rPPG framework. We believe that our contribution is conducive to practicalization and commercialization of the related research.
DOI:10.1109/ICBCB57893.2023.10246583