Adaptive filtering of high-resolution ECG signals

An adaptive filter for high-resolution electrocardiographic (ECG) signals is presented which estimates the deterministic component of the ECG signal and removes the noise. The filter needs two inputs: the signal (primary input) and an impulse correlated with the deterministic component (reference input). It uses the least-mean-square algorithm to adjust the weights in the adaptive process. A description of the characteristics and performance of this filter is presented. The adaptive was tested by simulation using real ECG signals and its performance was compared with the signal averaging technique. Several signal-to-noise ratios were considered and the effect of shape variations was studied. An application to ventricular late potentials is presented