Design Of Human Physiological Signal Acquisition And Analysis System

At present,people’s living standards have been greatly improved.However,unreasonable and unhealthy lifestyles pose a serious threat to people’s health.Human physiological signals contain a lot of pathological information.By detecting and analyzing physiological signals,dynamic monitoring of human physiological health status can be achieved.In order to explore the pathological information contained in physiological signals,this thesis designs a human physiological signal acquisition and analysis system,and proposes a deep learning algorithm based on pulse wave conduction time and photoplethysmography pulse wave based on knowledge driven,to achieve monitor users’ daily physiological health status.The main research content is as follows:(1)In response to the problems of existing physiological signal detection equipment is inconvenient to carry and expensive.A portable and cost-effective multisensor integrated acquisition terminal has been designed,which is mainly composed of pulse wave sensors,electrocardiogram sensors,blood oxygen sensors,body temperature sensors,and micro processing units.By picking up four physiological signals through sensors,the hidden physiological signals are visually displayed.By combining wireless transmission modules,the collection and transmission of physiological signals have been achieved.(2)In response to the weak low-frequency characteristics of physiological signals,the physiological signals collected by the integrated collection terminal are denoised and baseline drift removed using wavelet transform and median filtering,respectively;An adaptive threshold detection algorithm based on Hilbert transform is designed to locate the extreme points of signals,so as to complete the calculation of pulse wave transit time;Combined with the characteristics that the pulse wave transit time has a direct relationship with the changes of arterial blood pressure,based on knowledge driven,a deep learning algorithm based on pulse wave transit time and photoplethysmography pulse wave is proposed,which achieves high-precision acquisition of continuous blood pressure,the average absolute error of diastolic blood pressure and systolic blood pressure is 2.78 and 5.03 respectively,and the standard deviation is 5.23 and 6.36 respectively,indicating the effectiveness and reliability of the algorithm.(3)Based on the above research content,a human health monitoring system has been developed,mainly including program design for the front-end and server.The system has intelligence and convenience,with a mobile phone as the visual interface.On the mobile phone,it mainly completes user login,Bluetooth communication,data analysis,real-time waveform display,and network communication parts;The server is deployed in the local Tomcat,mainly completing the processing and analysis of physiological signal data,and returning the calculated blood pressure value to the frontend display.