ECG Signal Filtering Approach for Detection of P, QRS, T Waves and Complexes in Short Single-Lead Recording

Abstract

This article is devoted processing analysis measuring electrocardiographic information in conditions of high frequency and low frequency interference. These types of interference reduce accuracy and lead errors in the separation of informative components of the ECG signal. The article presents a new approach based on the criterion of minimizing measurement errors and selecting all informative components (morphologies) ECG signals. This approach based on choosing an optimal bandwidth bandpass filter based on this criterion and allows to increasing the accuracy of detection of variable waves P, T and QRS complexes. Effectiveness of using the optimal filter bandwidth is estimated quantiles confidence coefficient and correlation coefficient. Results of analysis show a statistically significant correlation selected optimal filter bandwidth with steady-state error for the processing analysis ECG signal. The efficiency proposed approach confirmed by results of computer simulation and experimental approbation. Computer simulation results show an increase in the quality selection of information components ECG signal. The filtered waveform showed identity with a given ECG signal from MIT database (Massachusetts Institute of Technology). The results of experimental approbation of this approach allow separating the variable waves P, T and QRS complexes on the background of real interference. Experiments have demonstrated improved accuracy of analysis processing and separation efficiency of the informative ECG signal components.