Research And Implementation Of Pulse Rate Variability Analysis System Based On Wrist-pulse Wave

At present,the routine physical health check-up in hospitals requires the placement of multi-connected electrodes in close parts of the body during the dynamic ECG HRV analysis.The collection process is complicated and restricts the daily activities of the users,and people are increasingly demanding the convenience and comfort of daily HRV detection.In recent years,noninvasive optoelectronic methods have attracted much attention for detecting tissue blood volume changes to obtain pathological features.A large number of experimental studies at home and abroad have shown that the PRV extracted from the variability analysis of PPG has high consistency with HRV.In this case,a PRV analysis system based on wrist-pulse wave is designed.The wearable acquisition device and the mobile terminal are connected wirelessly to realize real-time continuous analysis of the PRV parameters,and the auxiliary diagnosis of cardiovascular and mental stress diseases is conveniently and quickly performed.This thesis first describes the relationship between pulse wave and PRV.PRV is transformed into a sequence of time series waveforms by detecting small changes in the pulse waveform between beats.Then the pulse wave collection and the real-time transmission of the original PPG data are completed by the pulse wave collector.The STM32L151 is used as the central control unit to maintain the sampling rate of 250 Hz for A/D conversion to obtain the pulse value.The original pulse data is sent to the intelligent terminal via Bluetooth.Secondly,the pulse rate variability detection algorithm is studied.Pre-simulation and verification of PPG signal filtering algorithm and PRV time-frequency domain analysis method using MATLAB tool.The PPG signal filtering algorithm and the PRV time-frequency domain analysis method are simulated and verified by MATLAB tool.Five-point smoothing filter is selected to remove high-frequency interference of pulse signal.Differential point-by-point discriminant algorithm is used to identify P-point of pulse characteristic.The extracted PP interval sequence is analyzed by statistical analysis and FFT transform to obtain the time-frequency domain characteristics of pulse rate variability signal.Finally,PRV data were collected under different postures and intensity of exercise to verify the influencing factors of PRV parameters.The effectiveness and reliability of the entire PRV analysis system were evaluated,and experimental test specifications were developed.The time domain(SDNN,RMSSD)and frequency domain(TP,LF/HF)indicators were selected for 18 samples.The measurement results of the PC are standard values,the system terminal displays the results as measured values,and the two sets of data Bland-Altman consistent scatter plots are drawn.It is observed that more than 95% of the sample measurement differences fall within the consistency limit;The BAR values of time domain indicators(SDNN,RMSSD)are 0.089 and 0.063,respectively,and the BAR values of frequency domain indicators(TP,LF/HF)are 0.158 and 0.133,respectively,indicating that the design system can stably and accurately monitor the user PRV values.