Vital Sign Monitoring Using Ultra Wide Band Radar

Respiratory rate(RR)and heart rate(HR)are important components of vital sign,which are directly related to clinical diagnosis of patients’ physical condition.At present,the detection of RR and HR is mainly based on contact equipment,non-contact detection equipment is still in development.Ultra Wide Band(UWB)bioradar has the advantages of high ranging accuracy,low energy consumption and strong anti-interference ability.It has been developed rapidly in recent years and can be used for non-contact vital sign detection.Because UWB radar can be used for remote detection,which can effectively reduce the risk of infection of medical staff and has a positive effect on clinical monitoring.The main research contents and innovative work of this paper are as follows:(1)Construction of UWB radar system.The hardware design of radar varies according to different application scenarios.According to the indexes of hardware,UWB radar system for detecting physiological information is built in this paper.The system is harmless to human body and meets the requirements of use.(2)Study on respiratory signal extraction.The band-pass filter is used to denoise the echo signal,and the orthogonal matching pursuit(OMP)algorithm is applied to the respiratory signal reconstruction.The accuracy reaches 96.34%when the sparsity is 1.In addition,the OMP algorithm can compress data,and the compression efficiency is affected by the number of signals and the sparsity.When the number of signals is 10000 and the sparseness is 1,the storage space used is 3.13%of the original data.(3)Research on heartbeat signal extraction.In this paper,a data fusion algorithm based on signal characteristics is proposed.Differential Enhancement(DE)algorithm and Ensemble Empirical Mode Decomposition(EEMD)algorithm are used to process radar signals,and two heartbeat signals and corresponding signal characteristics are obtained.During the fusion,the signal characteristics are used to evaluate the signal quality.Different feature ranges correspond to different fusion methods.Compared with direct fusion,the proposed method can improve the accuracy by 4.52%.(4)Ultra wideband radar system verification.In order to verify the feasibility of the UWB radar system proposed in this paper,taking emotion recognition as an example,the radar was used as the front end to obtain respiratory signals and heartbeat signals,construct multiple features from time domain,frequency domain and nonlinear dimensions,and input them into the support vector machine to classify emotions.Under the cubic polynomial kernel,the recognition accuracy of three classification is 63.68%,which verifies the effectiveness of UWB radar system,and also demonstrates the reliability of signal processing method for extracting respiratory heartbeat.