Research On A Noninvasive Continuous Blood Pressure Monitoring Algorithm Based On ECG And Pulse Wave Signals

According to reports in recent years,the mortality rate of cardiovascular diseases is increasing year by year,and the demand for all-weather blood pressure monitoring is increasing day by day.The ability to quickly,timely and conveniently perform blood pressure monitoring has become a hot research topic.Noninvasive continuous blood pressure monitoring is a continuous blood pressure monitoring scheme that can achieve real-time all day without manual intervention.Noninvasive continuous blood pressure monitoring based on ECG and pulse wave is a hot research direction in recent years.Noninvasive continuous blood pressure monitoring based on ECG signal and pulse wave is to obtain the characteristic parameters of ECG signal and pulse wave signal,and use the method of linear regression analysis to establish the blood pressure equation model.The portability and continuity of this method make the whole day blood pressure monitoring easier and has a good application prospect.In this paper,a non-invasive continuous blood pressure monitoring algorithm based on ECG and pulse wave signal is studied.The method uses the characteristic parameters in the database to fit the blood pressure model,and designs a set of adaptive measurement algorithm for group measurement.Finally,we use the self-developed ECG and pulse wave signal acquisition equipment to collect signals,and demonstrate the feasibility of the whole algorithm through experimental analysis.Firstly,it is necessary to extract relevant characteristic parameters for fitting blood pressure model.Using the two lead ECG signal,pulse wave signal and arterial blood pressure signal in the database,the characteristic parameters extracted from these three signals can be used for multiple linear regression analysis,and a prediction model of blood pressure can be obtained.Secondly,the model is optimized and adjusted.Because the effect of blood pressure model prediction using all data fitting is not ideal,the data can be divided into three categories,which can reduce the error of model prediction.At the same time,an adaptive judgment method is added,so that the input characteristic parameters can be automatically determined to belong to the classification.Finally,we independently developed a Bluetooth platform for ECG and pulse wave signal acquisition,which transmits the collected signals to the upper computer through Bluetooth,and processes,extracts and calculates the data in the upper computer to get the estimated blood pressure.The results of the algorithm and the actual measurement results can meet the measurement standard of AAMI for electronic sphygmomanometer,which proves that the algorithm has a certain reliability.This study has undergone two sets of tests of static experiment and dynamic experiment.Under the static experiment test,the average absolute error with the commercial electronic sphygmomanometer is 4.05 mm Hg for systolic blood pressure and 3.91 mm Hg for diastolic blood pressure;The average absolute error of the sphygmomanometer is 7.71 mm Hg for systolic blood pressure and 4.63 mmHg for diastolic blood pressure.