Contact-free measurement of cardiopulmonary signatures via a microwave Doppler radar using adaptive filter

It has been proven that the respiration and heartbeat signals can be contact-free detected via a microwave Doppler radar. However, the heartbeat signal cannot be identified from the body-movement signal when the human subject breathes freely. To resolve the problem, the adaptive filter is used to simultaneously measure the heartbeat and the respiration signal. Firstly, the respiration and the bodymovement signal are filtered by finite impulse response (FIR) band-pass digital filter with the cutoff frequency of 0.03 and 3.3 Hz. Then, the respiration signal is regarded as reference input of adaptive noise canceller (ANC) and the body-movement signal is regarded as the primary input. Finally, the heartbeat signal is extracted from the body-movement signal using the recursive-least square (RLS) algorithm. Experimental results showed that the adaptive filter cannot only effectively remove the respiration signal, but also the heart rate derived from the filtered heartbeat signal was strongly correlated with that from the electrocardiogram (ECG) recordings (γ 2 = 0.95, p<0.0001, n=16).