Influence of High Pass Filtering on the T-wave End Estimation

The accurate estimation of T-wave end point is essential for the analysis of QT interval variability in ambulatory electrocardiogram, but it can be affected by high pass filtering (HPF) requested to eliminate the high baseline drifts (BLD) typical of this type of monitoring. In this work, the effect of Butterworth HPF on the T-wave end estimation is characterized using an algorithm based on the calculations of triangle’s areas. The evaluation was done using synthetic and real signals obtained from QTDB database. The results shows that for cutoff frequencies higher than 0,8 Hz, the error in the T-wave end estimation is inadmissible so is not convenient to use the linear HPF to reduce BLD, instead of, other type of filtering that doesn’t distort to the T wave.