Correction of False Diagnosis Recording in the Electrocardiograph Signal by Adaptive Digital Filter

Abstract

Electrocardiogram (ECG) instrument is used to provide diagnostic information about the critical condition of the patient’s heart, but its performance sometimes suffers from some sources of noise such as: 50Hz line frequency, Hum and high frequency (HF) noise. These sources of noise affect the performance of the ECG and hence cause false diagnosis recording that tricks the doctor who uses this machine. The current paper; proposes to tackle the problem of false diagnoses recordings by employing adaptive digital filter based least mean square (LMS) error algorithm in order to refine the ECG signals from the disturbing sources of noise. Based matlab programming; I have studied two different cases of mixing random noise that disturb the performance of the ECG instrument. The first kind of noise is mixing the line frequency 50 Hz with the ECG signal; the second kind of noise is mixing the hum and high frequency noise with the ECG signal. Numerical values for digital filter parameters have been used as: number of taps or order (M = 16), step size (μ = 0.005), sampling frequency (Fs = 1000Hz), interfering line frequency 50Hz, hum noise, and HF noise. The experimental results using Matlab simulation show that; the proposed scheme of employing digital filter based LMS algorithm; can tackle the problem of false diagnoses that causes frustration for the patient and tricks the doctor. The proposed scheme has several advantages such as; simplicity, reliability, practical applicability, adaptability to the change in signal characteristics and cost affordability.