Research On Compressed Sensing Reconstruction Of Pacing ECG

It is difficult to deal with the sampling,processing and transmission of a large number of pacing ECG for traditional pacemaker monitoring methods due to the restrict of signal frequency,power consumption,monitoring sites and other factors.Compressed Sensing theory could be sampling with low-power and high-efficiency,and recovering the reconstructed approximating the original using corresponding optimization algorithm in the monitoring terminal.This study is aimed at exploring a signal-aware and reconstruction algorithm for pacing ECG signals,and provide technical options for achieve pacing ECG remote monitoring system using compressed sensing technology.Combined with the theory of compressed sensing,this research aimed at the pacing ECG signal and researched the theoretical and experimental of the signal sampling and signal reconstruction in the process of monitoring.The main research work of the paper:Firstly,designed stochastic sparse binary measurement matrix which is suitable for the pacemaker ECG remote monitoring,and applied it in the signal perception and reconstruction.Secondly,based on the study of block sparse Bayesian learning,the compressed sensing of pacing ECG signal is studied to explore the signal reconstruction under the framework of the algorithm.Thirdly,a reconstruction algorithm suitable for pacing ECG signal is improved with the idea of biased reconstruction according to the characteristics of existing compressed sensing algorithms to resolve the problem of signal reconstruction in noise environment.Furthermore,the optimization method of the ridge parameters is studied to determine the optimal ridge parameter value of the pacing ECG signal reconstruction in the noisy environment.Fourthly,the improvement of the designed measurement matrix and reconstruction algorithm has been evaluated in research content.The perception efficiency and matrix structure has been compared for measurement matrix,and the reconstruction precision and generalization of adaptive capacity for noise has been compared for reconstruction algorithm.