Band-stop Filter Tuning for Power-line Interference Rejection in ECG Signals

2021 ◽  
Author(s):  
Georgy Mihov ◽  
Dimiter Badarov
Keyword(s):  
Power Line ◽  
Interference Rejection ◽  
Ecg Signals ◽  
Filter Tuning ◽  
Band Stop Filter ◽  
Power Line Interference ◽  
Line Interference

scholarly journals Power-line Interference Removal from High Sampled ECG Signals Using Modified Version of the Subtraction Procedure

2020 ◽  
Vol 24 (4) ◽  
pp. 381-392
Author(s):  
Ivan Dotsinsky ◽  
◽  
Todor Stoyanov ◽  
Georgy Mihov ◽  
◽  
...  
Keyword(s):  
Sampling Rate ◽  
Power Line ◽  
Signal Spectrum ◽  
Band Pass Filter ◽  
Pass Filter ◽  
Sample Number ◽  
Ecg Signals ◽  
Subtraction Procedure ◽  
Power Line Interference ◽  
Line Interference

The acquired ECG signals are often contaminated by residual Power-line Interference (PLI). A lot of methods, algorithms and techniques for PLI reduction have been published over the last few decades. The so called subtraction procedure is known to eliminate almost totally the interference without affecting the signal spectrum. The goal of our research was to develop a heuristic version of the procedure intended for ECG signals with high Sampling Rate (SR) up to 128 kHz. The PLI is extracted from the corrupted signal by technique similar to second order band-pass filter but with practically zero phase error. The sample number as well as the left and right parts outside the samples belonging to a current sine wave, which is extracted from the contaminated signal, are counted and measured. They are used to compensate the error arising with the shift between the moving averaged free of PLI signal samples and their real position along the linear segments (usually PQ and TP intervals having frequency band near to zero). The here calculated PLI components are appropriately interpolated to ‘clean’ the dynamically changed in amplitude and position contaminated samples within the non-linear segments (QRS complexes and high T waves). The reported version of the subtraction procedure is tested with 5 and 128 kHz sampled ECG signals. The maximum absolute error is about 20 μV except for the ends of the recordings. Finally, an approach to PLI elimination from paced ECG signals is proposed. It includes pace pulse extraction, signal re-sampling down to 4 kHz and subtraction procedure implementation followed by adding back the removed pace pulses.

Sign in / Sign up

Export Citation Format

Share Document