QT interval time frequency analysis using Haar wavelet

Since the wide variety of changes observed in the morphology of T wave and its low frequency components, QT interval measurement algorithms have been an intensive research field in the recent years. Also, the observed changes in T wave morphology at different heart rates have made the QT Interval measurement a difficult task in exercise ECG analysis when repolarization time parameters are desired. The present study applied discrete wavelet transform (DWT) analysis employing the Haar wavelet in the exercise ECG signal. DWT using Haar wavelet in second dilation was obtained from every ten seconds of the exercise test, then a envelope detection procedure was applied over the previous signal. An algorithm to correlate the envelope detection and the zero crossing of the obtained wavelet was applied to set the QRS complex beginning, T wave peak and the T wave end. Comparing measurements of the present methodolgy with those obtained by expert clinical view shows a very similar performance. The authors´ analysis was applied in signal obtained from a set of 30 coronary artery diseased patients and 28 healthy volunteers