Research On Vascular Health Assessment Method Based On Blood Volume Pulse Wave And Design Of Network System

With the progress of the society,the improvement of living standards have raised.At the same time,life pressure,unreasonable diet and so on have brought the potential crisis of cardiovascular disease.If we can understand the vascular health problems and prevent them in advance,it is of great significance to the prevention of cardiovascular diseases.The formation of pulse condition is not only related to heart,blood and pulse,but also closely related to other visceral functions.This paper is based on the vascular volume pulse wave to complete the identification and classification of pulse,so as to further evaluate the vascular health.Firstly,a pseudo-error elimination algorithm based on the characteristic K value is designed for the photoelectric volumetric pulse wave collected from the hardware front end,and the signal is preprocessed.The K value of photoelectric volumetric pulse wave signal is calculated,and the time domain recursive method is used to test whether the pulse wave has false error interference.If there is false error interference,the estimated value is used to replace it,and then the false error interference caused by environment and other factors is eliminated.The relative error of K value estimated by this method is 0.82%through experimental checking,and the algorithm is reliable.Then,in order to quantify the clinical pulse signal objectively,the pulse wave signal was extracted in the time domain.The sliding window method was used to find the peak and valley points of pulse wave and divide each segment.Four characteristic values were extracted preliminarily in the time domain.In order to ensure the completeness of feature extraction,the Hilbert yellow transform is used to process the signal,and the energy value of its marginal spectrum is calculated in the frequency domain,and each frequency band is divided according to frequency.Through the analysis and calculation of 200 cases of pulse signal,it is found that the energy value in frequency domain of pulse wave signal of people with abnormal pulse pattern is generally concentrated in a certain frequency band,with a standard deviation of 19.04,while that of normal people is evenly distributed,with a standard deviation of 6.02.There is a difference between the two,and the extraction of eigenvalue is reliable.According to the parameters of photoelectric volumetric pulse wave,BP neural network,LM algorithm optimized neural network and genetic algorithm combined neural network were used to classify and recognize pulse signals.Through the comparison of the recognition results,it is found that the BP neural network based on genetic algorithm has a higher recognition rate than the previous two models,and the accuracy rate reaches 95%,which better realizes the classification and recognition of pulse signals.In order to realize the practical application of the algorithm,the hardware acquisition,algorithm and front-end interface display are integrated into the cloud platform system.The cloud platform described in this paper is based on the Web application of Spring,Spring MVC and MySQL database,and Ali cloud server is used to build a networked system.Real-time pulse wave acquisition waveform is displayed in detail.Users’ personal information as well as physiological parameters,health assessments,etc.The application has been running normally in Aliyun server,and users can view the signal and blood vessel assessment status in real time through the web page,which has solved the original needs of mobile and networked system.