Non-harmonic Analysis Based Instantaneous Heart Rate Estimation from Photoplethysmography
Fangyu Li, Wen-Zhan Song, Changwei Li, Aiying Yang
Abstract
Instantaneous Heart Rate (IHR) detection is important but challenging. As a kind of biomedical signal, photoplethysmography (PPG) is a good source to extract the intrinsic IHR information for further diagnostic interpretation. However, because of the non-stationary characteristics, the traditional IHR estimation results can be unreliable, and may be contaminated with artifacts and noises. Even state-of-art time-frequency analysis techniques can not fully handle the subharmonics interferences to obtain a reliable and robust IHR estimation. In this paper, we propose a hybrid IHR estimation approach based on the non-harmonic analysis (NHA) and an advanced time-periodic transform (TPT). We propose to use a NHA model to adaptively extract intrinsic oscillatory modes from PPG data. The previously unsuppressed sub-harmonic components can be removed thanks to the extracted mono-frequency components. Then the IHR can be estimated from the left ridge on TPT. Experiment results demonstrate the advantages of the proposed approach, which is promising for PPG signal processing and analysis.
BibTeX
@inproceedings{icassp2019_nonharmonicanaly,
title = {Non-harmonic Analysis Based Instantaneous Heart Rate Estimation from Photoplethysmography},
author = {Fangyu Li and Wen-Zhan Song and Changwei Li and Aiying Yang},
booktitle = {ICASSP 2019},
year = {2019}
}