Electrocardiogram signal processing by morphological operators

A new approach to impulsive noise suppression and background normalization for digitized electrocardiogram (ECG) signals is presented. The algorithm makes use of a class of nonlinear signal-processing operators known as morphological operators for processing the ECG data. A brief introduction to these nonlinear self-processing operators and a detailed description of the algorithm are presented. The performance of the algorithm is evaluated by test data made up of a known ECG signal, digitized from an analog simulator and corrupted by impulsive noise and background drift, and by processing clinically required ECG data from a standard set. It is concluded that the algorithm is effective for impulsive noise suppression and background normalization