Research And Implementation Of Key Algorithms For ECG Signal Processing

In our country,heart disease is the biggest threat to people's health.Real-time ECG monitoring of patients with cardiovascular diseases using Holter monitor is a convenient and effective method to reduce morbidity and mortality in patients with heart disease,especially for the diagnosis of arrhythmia and prediction of sudden cardiac events.In order to be more accurate and convenient for early warning and diagnosis of arrhythmia,this thesis studies and realizes the key ECG signal processing algorithm for real-time Holter monitoring.The main research contents include the following points:1.Research and implementation of arrhythmia detection:The differential threshold algorithm and wavelet transform algorithm are used to identify QRS wave respectively.And different arrhythmia were diagnosed based on QRS wave feature extraction.Using the MIT-BIH database samples to verify two algorithm's feasibility and accuracy,and to compare the performance of the two algorithms.2.Research and implementation of 12-lead ECG reconstruction algorithm by 3-lead ECG:The pre-processed lead ECG signal and ECG monitoring data are trained through the LM artificial neural network,to get the reconstruction model of 12-lead ECG.Finally,the correlation and the root-mean-square error between the reconstructed and the original waveform are used to compare the effects.3.Research and implementation of ECG compression algorithm:ECG signals are compressed by AZTEC algorithm and wavelet transform algorithm respectively,and the performance is analyzed and compared from the aspect of compression ratio and speed.4.Design of 3-lead ECG electrode connection discrimination algorithm:The 3-channel standard ECG signals are trained by the feedforward neural network algorithm to obtain a discriminant model of the electrode lead error,which can effectively determine whether the ECG electrode is connected to the wrong electrode or not.