Independent effects of low frequency magnitude and phase distortion on the recorded electrocardiogram

Abstract

Electrocardiographic signals must go through electronic amplification and filtering to increase amplitude and to diminish noise before they are recorded. However, amplification and filtering may introduce unwanted magnitude and phase distortion in the recorded signal. The relative importance of inadequate magnitude versus inadequate phase responses of electrocardiographic recording systems has been a subject of debate. Although electrocardiographic signal distortion due to inadequate magnitude response has been identified, 1 there is evidence 2,3 that phase distortion has a greater effect on clinical interpretation of the recorded electrocardiogram. This study was designed to quantify the independent effects of low-frequency magnitude and phase distortion on electrocardiographic signals. Electrocardiographic signals from Holter recordings in 10 patients were digitized and processed by magnitude-modifying and phase-modifying filters. The independent effects of magnitude and phase distortion were evaluated by calculating cross-correlation coefficients between filtered and unfiltered waveforms in each patient.