Research On Continuous Blood Pressure Measurement Based On Pulse Wave

Blood Pressure(BP)is the blood pressure in the artery to per unit area of the Vascular wall,it is an important physiological parameter that reflects the function of the cardiovascular system of the human body.It is an important basis for clinical diagnosis and family health care,including Systolic Blood Pressure(SBP)and Diastolic Blood Pressure(DBP).Effective measurement of blood pressure has important implications for the prevention and treatment of cardiovascular diseases.Continuous Blood Pressure measurement can measure BP in each cardiac cycle,which means it can monitor the change of BP in real time.However,the current clinical measurement of Blood Pressure is still based on intermittent measurement.That is only a single measurement for a moment,and cannot get rid of the shackles of the inflatable cuff.The measured values of Blood Pressure can not accurately reflect the patient’s Continuous Blood Pressure level in a certain period of time.,Therefore,it is very important to study the method of Continuous Blood Pressure measurement.Under such a background,this paper on the basis of Arterial Blood Pressure and pulse wave propagation theory,taking the pulse wave for the main object of study,to find the relationship between the pulse wave and BP,and study the method based on support vector machine(SVM)to realize continuous BP measurement.This research is divided into two parts:The first part is the pulse wave denoising and feature points detection.In this paper,the BP and pulse data of the same heart beat are mostly from the MIMIC database,also,we collect these data ourselves for the research.The pulse wave data were pre processed,and then the feature points were extracted.For the removal of the noise in the pulse wave signal,the wavelet based method is adopted to remove the baseline drift;And use digital filter to filter out high frequency interference.For the detection of characteristic points of pulse wave,a algorithm which based on differential threshold was used,Finally,the feature points like pulse wave onsets and peak points are extracted accurately.The second part is the research and analysis of Continuous BP measurement method based on SVM regression.Take the feature points detected before as reference points to divide the pulse wave in different ways,it’s frequency characteristics were also considered in this research.Select different pulse wave characteristics as the input of the model,and take the SBP or DBP of the same heart beat as the target output.Then,different train sets and test sets can be established.Use these sets to train BP Regression Model,and then predict BP values beat to beat with the model.Through the comparison of different experimental schemes,the results show that,when taking pulse wave onsets as reference points to divide the pulse wave,and taking the single-cycle sequence of the pulse wave as input feature,the SBP or DBP of the same heart beat as the target output,then the BP Regression Model obtained can measure BP values better.Experimental results show that,the average error of SBP and DBP measurement are all less than 5mmHg,and the standard error are all less than 8mmHg,this can meet the standards of the Association for the Advancement of Medical Instrumentation(AAMI).