Why the study?
In vital sign radar technology, respiration harmonics distort and overwhelm heartbeat measurement due to large signal amplitude, presenting signal processing challenges.
A supervised machine learning gamma filter can effectively separate heartbeat signals from respiration artifacts in Doppler radar measurements in real time.
May enhance radar heartbeat extraction amid respiration harmonics; leaves open clinical validation before practice change.
The advancement of vital sign radar technology has proven to be a useful tool in assessing various physiological dynamics, including heartbeat and respiration. There remains several signal processing challenges in this field, which include overcoming the nonlinearities and harmonics that populate the power spectrum. Respiration harmonics distort and overwhelm the measurement of heartbeat due to the large signal amplitude. A supervised machine learning algorithm, the gamma filter, offers an efficient, calibration-free solution to model the time series heartbeat signal given respiration and respiration artifacts. The measured signal is provided by a 5.8-GHz quadrature Doppler radar and a modified electrocardiogram signal is used as the ground truth for training the filter. Experimental results show that the heartbeat is independent and separable from respiration and the algorithm can be implemented in real time.
No takes yet. Share an insight, caveat, or question.
Saluja et al. (2019) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: