Design Of All-lead ECG Monitoring System With Fall Function Detection

Nowadays,cardiovascular diseases and falls are two major factors that endanger the health of the elderly.With the appearance of aging problems in our country,the monitoring of physiological status and behavior of the elderly has important research value.If it can be found for the first time that an elderly person has an abnormal electrocardiogram or falls,it is crucial for timely medical treatment and saving the lives of patients.However,the vast majority of current portable ECG monitoring systems do not have the ability to judge falls.Therefore,it is very important to design a portable monitoring system with falls judgment and abnormal electrocardiogram analysis.Firstly,this article introduces the hardware design of the multi-functional monitor system.In the front of the analog circuit,the ADS1298 chip designed by TI which especially collect human biological electrical signals is chosen to collect the human body's ECG signals,the MPU6050 chip which is a 6-axis motion processing module is chosen to collect human body's acceleration value.In the digital circuit part,the central processing chip of STM32F103 series is selected,and all datas are transmitted to the host computer through Bluetooth.Secondly,there are many interferences in the ECG signal,at the first,50 Hz adaptive noise filtering method is used to remove the 50 Hz power frequency interference noise,and then the wavelet decomposition is used to decompose the ECG signal in 8scales,by removing the low frequency wavelet coefficients and heights,the information corresponding to the frequency wavelet coefficients removes baseline motion noise and high frequency noise interference.Finally,the support vector machine algorithm is used to classify the heart rate anomaly detection,and then the particle swarm optimization algorithm is used to optimize the parameters of the support vector machine,we obtain the optimal support vector machine parameters,and finally use the data in the MIT ECG database to test.The classification results show that the SVM classification algorithm optimized by parameters has a good effect and the classification accuracy is high.At the same time,compared with the classification results of BP neural network algorithm,the results show that the classification result of SVM algorithm in this paper is obviously better than BP neural network algorithm.In terms of the fall judgment algorithm,an algorithm based on threshold judgment is adopted.By taking the combined acceleration,the mean value of the acceleration and the human inclination as characteristic quantities,a reasonable threshold value is set to judge the fall.Finally,the volunteer is invited to perform a fall experiment,experiment results also prove the feasibility of the fall judgment algorithm.This paper also proposes the logical judgment relationship of merging heart rate abnormality and fall judgment,it assists in judging heart rate abnormality according to the result of the fall judgment,and carries out different levels of alarm according to the severity of the diagnosis result.