Research And Design Of ECG Early Warning System For Edge Of Network

Cardiovascular disease has a high morbidity and mortality rate.It has a higher death rate than other diseases in China.So the harm of cardiovascular diseases is great.Because of fast pace of life,increasing pressure of work,chaotic schedule and other reasons,the incidence of cardiovascular disease is rising rapidly.So it is very important to find effective cardiovascular disease prevention and treatment methods.Traditional ECG examination can accurately reflect the activity process of the heart,which can obtain rich pathological information and diagnose common cardiovascular diseases.However,traditional ECG examination cannot provide real-time detection,and cannot provide long-term ECG detection and abnormal warning for people in life.Therefore,this topic designs a wearable ECG early warning system based on the application background of network edge.The T-shirt was used as the wearing carrier of the acquisition and processing device.USB OTG data line is used as the differential lead line,and the differential lead line is used to connect the conductive silica gel electrode and the acquisition and processing device.Then the signal acquisition chip and the main control processing chip on the device are used to complete the digitization,processing and analysis of ECG signals.Finally,the physiological parameters obtained from the treatment and analysis were used to evaluate the health status and give early warning in case of abnormal ECG.Baseline drift,power frequency interference and EMG interference are common ECG noises.These noises will interfere the feature extraction of ECG signals.For this reason,the pre-processing algorithm is studied and compared in this thesis,and choose small calculation,easy to implement morphological filtering and smooth filtering to remove noise.After the noise removal,the difference threshold method with high real-time performance,small computation and easy realization was selected to complete the QRS localization of ECG signals.Aiming at the shortcomings of the traditional algorithm,the difference threshold method is improved so that the acquisition and processing device can output the ECG signal and its signal characteristics stably.In order to realize the ECG warning system at the edge of the network,five common ECG anomalies are studied in this thesis.This thesis get the ECG abnormal warning scheme by analyzing the difference of abnormal ECG signals,so as to realize the ECG abnormal warning.In addition,the system will be implemented through the way of software processing mode,sending mode and standby mode three states.The warning value set by the system is used to realize the direct conversion of the three modes.In this way,the power consumption is reduced and the stability of ECG early warning system is improvedFinally,in terms of the device,the experiment proves that the system has small size,low power consumption and long battery life.The preprocessing algorithm and feature extraction algorithm were tested by MATLAB simulation test and real machine test.In terms of algorithm,the test was carried out through Matlab simulation test and real machine test.Experimental results show that the preprocessing algorithm has a good denoising effect.And the feature extraction algorithm can accurately locate the R wave,the average accuracy of the algorithm is 99.24%.To sum up,the device is easy to carry,low power consumption,long battery life,stable output signal and other characteristics,each index to meet the relevant test indicators.