Research On The Non-contact Vital Sign Detection Radar And Its Physiological Information Processing And Recognition Technology

Non-contact vital sign detection radar technique is a hot research field in both biomedical sensing and modern radar,which has been rapidly developed in the past two decades.Some developed countries such as USA,Germany and Japan attach great importance to developing this technique.Some institutes and companies have been devoting to promoting this technique for both clinical medicine and home healthcare.Compared with the conventional contact vital sign sensors,this technique has the advantages of non-contact measurement,high sensitivity and high accuracy,which is especially suitable for long-term monitoring and special patient(e.g.burned patient)monitoring.Compared with other non-contact sensors(e.g.camera and acoustic sensors),this technique has the advantages of high adaptively,high penetrability and no privacy problems.Nowadays,there are several problems in the researches of the non-contact vital sign detection radar.First,the imperfection exists in the hardware of the non-contact vital sign detection radar,the DC offset in the quadrature channels affects the dynamic range and accuracy of vital sign signal.Second,the existing vital sign extraction algorithms are limited by the time resolution,which can not reveal the time-varying characteristics of the vital sign signals.Lastly,recent researches on this technique focus on the accurate measurement of respiratory and heartbeat frequencies and do not further investigate the intrinsic information hidden in the time-domain vital sign waveforms.Therefore,how to design the optimization algorithm for the DC offset in the quadrature channels,how to develop the instantaneous vital sign extraction algorithm for non-contact vital sign detection radar,and how to acquire more intrinsic information via the non-contact vital sign detection radar,are the research focuses of the non-contact vital sign detection radar.Based on these points,the contributions of this thesis are shown below:1.The research of the DC offset calibration algorithm for baseband signals in the non-contact vital sign detection radarThe null/optimal detection points problem exists in the quadrature receiver of the non-contact vital sign detection radar,which may reduces the dynamic range of the demodulated vital sign signal.To solve this problem,a DC offset calibration method based on nonconvex minimization has been proposed.In the proposed method,the DC offset calibration problem has been modeled using the l_p(0<p<1)minimization,which can be solved by the IRL1 algorithm to estimate the accurate DC offset.Compared with the results without DC offset calibration,the SNR of the demodulated signal are significantly improved.Compared with the results using l₂and l₁ minimizations,the accuracy of DC offset calibration can be obviously improved.2.The research of the instantaneous vital sign extraction algorithm based on the non-contact vital sign detection radarAs we know,the vital sign signals(e.g.breathing and heartbeat)are time-varying and non-stationary.The conventional algorithms for vital sign detection radar are mainly based on Fourier transform.These algorithms do not have time resolution,which can not extract the time-varying vital sign parameters.In this thesis,a novel instantaneous vital sign extraction algorithm based on the Synchrosqueezing transform has been proposed.The time-frequency spectrum derived from the Wavelet transform is squeezed along the frequency axis so that the high-resolution time-frequency distribution can be obtained.The experimental results demonstrate that the proposed algorithm can estimate the instantaneous breathing and heartbeat rates,which has some reference for researching the physiological dynamics.3.The research of the breathing disorder recognition system based on the non-contact vital sign detection radarRecent research on the breathing detection using the vital sign detection radar focuses on the measurement of the breathing rate,which does not further investigate potential vital sign parameters.In this thesis,a breathing disorder recognition system based on the digital-IF vital sign detection radar has been proposed.Through analyzing the relationship between multi-samples,the mapping model of breathing waveform-breathing pattern-breathing disorder has been constructed,so that the long-term breathing pattern monitoring and breathing disorder recognition can be realized.Both laboratorial and experimental results show that the proposed system can achieve long-term breathing disorder recognition with high robustness.4.The research of the beat-to-beat blood pressure monitor based on the non-contact vital sign detection radarRecent research on the heartbeat detection using the vital sign detection radar focuses on the measurement of the heart rate.In this thesis,a non-contact beat-to-beat blood pressure measurement technique has been proposed.Through the radar-detected abdominal aortic pulse wave,the important points are extracted so that the pulse transit time can be acquired.Using the mapping relationship between blood pressure and pulse transit time,the beat-to-beat blood pressure can be calculated,along with six kinds of parameters to evaluate the function of cardiovascular.The experimental results show that the proposed system can obtain the beat-to-beat blood pressure,along with the preliminary evaluation for cardiovascular system when compared with the electronic sphygmomanometer.