Evaluation of a damping coefficient influence made by notch filters on efficiency of ECG signals processing

Introduction. The aim of this study is to evaluate the effect of damping coefficient ξ provided by narrow-band notch filters on the quality indicators of the ECG signal processing accompanied by development of a special software package for experimental testing of the cascade filtering system of natural electrical noise and its harmonics. Materials and Methods. For the formation of additive components of signal and noise, test ECG signals from the Massachusetts Institute of Technology were selected as a model, and electrical noise was synthesized taking into account their harmonic nature. The transfer function of the notch filtering system was tuned and calculated. Quantitative quality metrics was used to determine the processing efficiency at the input and output of the signal filtering system. To assess the effect of damping coefficients made on the ECG signal processing system, the method of the linear regression analysis was applied. Methods of structural and procedural programming were used to create a software package for processing a full-scale noisy signal recording. Results. An increase in the efficiency of the ECG signal processing system was revealed when calculating quantitative indicators characterizing the quality solution to the signal filtering problem. It has been found that at values of the damping coefficient ξ = 0.1, the efficiency of the analysis of ECG signal processing in terms of accuracy and noise resistance increases, and the filtering error decreases due to the improvement of the selectivity of the notch filter at the resonant frequency. The effect of the damping coefficient on the quality indicators of ECG signal processing was established, and regression equations were obtained to characterize the adequacy of the designed model for the confidence interval (P = 0.99) at p <0.01. The performance of the developed software package has been demonstrated, which combines a cascade filtering system with a damping coefficient ξ = 0.1, to eliminate the natural electrical interference of the ECG signal and its harmonics. Conclusion. The evidence data obtained in our work, both calculated (theoretical) and practical (experimental) data, confirm the efficiency of the proposed processing approach to the selection of informative components ECG signal under the influence of high-frequency electrical interference and its harmonics. The sequential incorporation of two narrow-band notch filters with a damping coefficient ξ = 0.1 to suppress interference at a frequency of 50 Hz and its harmonics at a frequency of 100 Hz opens up new possibilities for processing ECG signals.