Research And Application Of Boundary Detection Method Of ECG Signal

Cardiovascular disease is the number one killer of human health,and it has become a major public health problem.The prevention and treatment of cardiovascular disease is an urgent task.Electrocardiogram(ECG)is the most convenient and effective method in the diagnosis of many cardiovascular diseases,and the peaks and boundary points are important features for diagnosis.The boundary points in the ECG present the state of the heart and are closely related to human health.In this paper,in order to realize the ECG signal reference point detection,the ECG signal boundary detection method is studied;The current status of wearable ECG devices on the market is analyzed,the deficiencies of existing devices are improved,and clothing of twelve-lead ECG monitoring is designed;the algorithm is verified by simulation experiments and multiple data sets,and the actual performance is better.The main research contents of the paper are as follows:1.Aiming at the problem of real-time detection of ECG data,a QRS detection algorithm based on adaptive threshold and Particle Swarm Optimization(PSO)is proposed.Aiming at the influence of motion artifacts,a QRS enhancement technique based on RS negative slope and duration is proposed.The algorithm completes the preprocessing through filtering,introduces the RS part in the signal enhancement stage,and utilizes the slope characteristics of the RS wave to achieve artifact suppression and improve the robustness of the algorithm.The decision-making stage combines two sets of adaptive thresholds to implement QRS detection,and uses PSO to select the best parameters.Finally,based on the peak value of QRS,the signal mutation is used to realize the boundary detection.This method reduces the amount of calculation,improves accuracy and robustness,and solves the problem of ECG real-time detection and the effects of motion artifacts.2.For the boundary detection of P and T waves,this paper proposes a detection algorithm based on wavelet transform,which introduces SASS(Sparsity Assisted Signal Smoothing)in the data preprocessing stage.In the preprocessing stage,SASS is used to process the data,which improves the destruction of the signal itself by low-pass filtering,improves the robustness of the algorithm,and restores more realistic ECG data.The boundary detection part uses wavelet transform to perform corresponding window processing according to the position of QRS,and finally realizes the multi-form P and T wave boundary detection.3.In view of the shortcomings of existing wearable devices,clothing of twelve-lead electrocardiogram monitor was designed,combined with the proposed algorithm,and finally realized the application.The clothing realizes the real-time collection of twelvelead dynamic electrocardiogram data,and transmits it to smart devices through Bluetooth.Combined with the cloud platform,the interaction between patients and doctors is realized.The clothing uses copper-nickel fabric material as the sensing electrode,which realizes efficient and comfortable data sensing.Combined with the proposed boundary detection algorithm,the electrocardiographic data is effectively processed,and the ECG signal collection,processing and application are realized.