.png){kind=icon}
.png){kind=icon}
Return
Combining blind source separation and adaptive filtering for motion artifacts removal from Photoplethysmography in different states
.png){kind=icon}
PRE
AI
X
C
G
Z
L
DOI：10.1016/j.bspc.2026.111122.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Photoplethysmography (PPG) signals have important applications in physiological parameter monitoring for wearable devices, but they are susceptible to motion artifacts (MAs) interference during exercise, leading to signal distortion and inaccurate heart rate monitoring. Aiming at the limitations of traditional methods to suppress motion artifacts in dynamic scenes, this study proposes a cascade denoising method combining blind source separation (BSS) and adaptive filtering. The method uses the accelerometer signal as the a priori information, initially separates the PPG signal from the motion artifacts by the BSS algorithm, and then uses the separated interference components as the reference signal, and further suppresses the residual noise by Volterra adaptive filtering. The experiments included PPG signals and triaxial accelerometer data from eight subjects in four exercise states, and the performance of the method was evaluated by the signal-to-noise ratio (SNR), pulse waveform index (PWI), and the energy share of heart rate band. The results show that compared with the traditional method, the proposed method significantly improves the quality of the PPG signal, effectively improves the PWI, makes the PPG morphology more regular, and the energy share of the heart rate band reaches 94.5%, and exhibits good robustness under different exercise intensities. This study provides an effective solution for accurate physiological monitoring of wearable devices in dynamic environments.
Journal
IF:
4.9
Papers:
9.7K
Citations:
2.4W
Organization
Cited Papers