Detection and Parameter Estimation of R Peaks in ECG Signal Using Optimization Algorithm

Abstract

Accurate R-peak detection is very important for arrhythmia diagnosis. Our previous effective R detection algorithm consisted of three strategies: band-pass filter, adaptive definition of interesting block and dynamic threshold. Then, it adopted the optimization algorithm to replace the knowledge-based theory and found out the suitable parameters (F1, F2, N, W1, W2, β and µ) in R detection algorithm quickly and obtained the high performance of detecting R peaks (99.77%). In order to improve the performance of the previous study, this study proposes to add the median filter in the algorithm to correct baseline wander components of electrocardiography (ECG) signals. It is necessary to defined two parameters (T1 and T2) in median filter. Therefore, this study adopts particle swarm optimization (PSO) to find the suitable parameters (T1, T2, F1, F2, N, W1, W2, β and µ) in the proposed method. The proposed method is applied to MIT-BIH arrhythmia database. The results show that PSO can find out the suitable parameters in R detection algorithm and have a higher accuracy (99.95%) than one of the previous study.