Research On Dynamic ECG Monitoring System For Automatic Diagnosis Of Arrhythmia

Arrhythmia refers to a problem with the frequency or rhythm of the heartbeat,and severe arrhythmia may cause lethal heart disease.Electrocardiogram(ECG)is a technique that uses to record changes in the electrical activity of the heart during each cardiac cycle.Clinically,doctors usually diagnose arrhythmia by analyzing the patient’s ECG with his/her relevant medical history and clinical manifestations.However,abnormal ECG signals usually occur by chance,which cannot be obtained from a short-term ECG.It is time-consuming and labor-intensive and lacks objectivity to only rely on manual processing of a patient’s long-term ECG records.Moreover,centralized analysis after recording ECG lacks real-time performance and cannot well deal with the sudden risk of patients.With the development of technology,Domestic and foreign research have made good progress in the field of ECG acquisition and arrhythmia classification algorithms.However,there are relatively few studies on the integration of "ECG acquisition + ECG diagnosis + remote management".Therefore,it is very meaningful to study a Holter monitoring system with cost effective,high reliability,automatic diagnosis of arrhythmia and remote data management functions for individuals,families,communities,and hospitals.On the one hand,it can help users prevent sudden risks,and on the other hand,it can help effectively save medical resources.This article mainly studies the Holter monitoring system for real-time diagnosis of arrhythmia.The work mainly includes the following parts.1.Designed and developed a Dynamic ECG data collector.This collector takes STM32F405 RG as the core,adopts the 24-bit integrated analog front end ADS1291 to design the signal conditioning module,which can acquire the user’s ECG signal in real time,transmit it to the ECG monitoring client in a wireless way,or save it to the TF memory card.The collector has the technical characteristics of low power consumption,miniaturization,and wearable.2.Designed and developed ECG monitoring client software and ECG data remote recording management system.The ECG monitoring client software is developed based on Python’s Py QT5 interface library.It has functions such as ECG preprocessing,arrhythmia diagnosis,visualization,and features of flexible deployment.The ECG data remote record management system provides a RESTful interface to upload client data under the Django framework,and design a data management website.3.Studied the real-time arrhythmia diagnosis method based on deep neural network.First,the forward feedback neural network model for real-time QRS complex detection is designed,and the ECG time domain features are extracted from the R-wave position to construct a one-dimensional convolutional neural network model for real-time arrhythmia diagnosis.Finally,the MIT-BIH arrhythmia database was used to train the model and test the diagnosis algorithm.The results show that the algorithm has a recall rate of 98.0%,a precision rate of 99.5%,and an overall accuracy rate of 97.6% for the cross-patient QRS complex position detection,and the accuracy rate for the 5-class arrhythmia classification is 91.5%.Finally,the function of this dynamic ECG monitoring system is verified.The results show that the dynamic ECG monitoring system studied in this thesis,which is used for automatic diagnosis of arrhythmia,has the functions of real-time ECG acquisition,wireless transmission,real-time QRS complex detection and remote ECG data management,and meets the design requirements.