The Effects of 0.67 Hz High-pass Filtering on the Spatial QRS-T Angle

Abstract

The spatial QRS-T angle (SA) is a vectorcardiographic (VCG) parameter that has been identified as a marker for changes in the ventricular depolarization and repolarization sequence. Recent research has developed a number of different linear electrocardiographic lead transformation matrices that have removed the lead system related barriers for the utilization of the SA in monitoring applications. However, monitoring applications utilize upon high-pass filters with cutoff frequencies as high as 0.67 Hz. Such high-pass filters are known to influence the ST-segment of the electrocardiogram (ECG). The aim of this research was to quantify the influence of different 0.67 Hz high-pass monitoring filters on the SA. We assessed the differences between the SA values computed from 0.05 Hz high-pass filtered Frank VCGs and SA value obtained from different 0.67 Hz high-pass filtered VCGs (0.67 Hz zero-phase high-pass filtered Frank VCGs, 0.67 Hz nonlinear-phase high-pass filtered Frank VCGs and 0.67 Hz zero-phase high-pass filtered VCGs derived from the Mason-Likar (ML) 12-lead ECG). The differences between the SA values were determined using a study population of 181 subjects and quantified as random error (span of Bland-Altman 95% limits of agreement). The random error magnitude due to the derivation of the VCG from ML 12-lead ECG data was not found to be substantially increased by the utilization of the zero-phase 0.67 Hz high-pass filter (random error magnitude based upon ML derived 0.67 Hz filtered VCG data: 45.40° vs. random error magnitude based upon ML derived 0.05 Hz filtered VCG data 39.57°). This finding suggests that it is possible to utilize zero-phase (or alternatively linear-phase) high-pass filters in SA monitoring applications that utilize VCG data derived from the ML 12-lead ECG.