Design Of Respiratory Rate Extraction System Based On PPG Signal

As an important factor in evaluating the physiological state of the human body,Respiratory rate can provide some important information about human health.At present,the clinical measurement methods are mainly pressure sensor method or spirometer method,which have a certain compulsion.With the development of non-invasive detection technology,photoplethysmography(PPG)has been widely used in daily life because of its non-invasiveness and reliability.People are gradually paying attention to the early warning of respiratory diseases,so it is of great significance to carry out accurate and effective respiratory signal monitoring.Taking respiratory rate monitoring as the research background,this paper proposes a method to extract and reconstruct respiratory signal from pulse wave,and calculate respiratory rate.The main work is to design the micro system hardware with low power consumption,build the software part of the lower computer,design the interface of the upper computer,write the program of the lower computer,and propose a new algorithm.The hardware design of the low-power micro-system has accomplished the detected function of human physiological parameters.The lower computer part was used the MAX30102 sensor to realize the function of PPG signal detection.The HKH-11 B abdominal pressure sensor was used to measure the respiratory signal during the same period as the reference respiratory signal.The host computer has used the WPF(Windows Presentation Foundation)technology to build a platform to display the pulse and the respiration signals collected in real time.According to the non-stationary and nonlinear characteristics of the photoplethysmography,the design of the algorithm adopts the method of Ensemble Empirical Mode Decomposition(EEMD)combined with wavelet threshold to test the PPG signal in the MIMIC database and selects the components with high correlation to reconstructed the breathing signal.The feasibility of the algorithm is verified.After that,the PPG signal collected by the hardware system in this paper is analyzed and processed,the respiration signal obtained after reconstruction and the reference respiration signal are subjected to waveform relative coherence analysis,AR power spectral density estimation,and respiration rate is realized by combining the time-frequency domain.The experimental results show that the relative coherence coefficient of the reconstructed respiratory waveform is above 0.62,the power spectral density correlation coefficient of the AR model is above 0.97,the accuracy of the estimation of the respiratory frequency is above 95%.This system simplifies the way of monitoring respiratory signal at home,verifies the feasibility of EEMD-wavelet threshold algorithm to extract respiratory signals from PPG signals and the reliability of hardware system to monitor breathing signals and achieves the expected goal.Combined with the relevant parameters of the respiratory signal,the system can evaluate the health index of the human respiratory system,prompt early warning of possible respiratory diseases.