Dynamic measurement of the QT interval

Algorithms for automatically estimating the R, Q, and T fiducial points have been developed, and their sensitivity to baseline noise and wave morphology fluctuations has been tested on simulated ECGs (electrocardiograms). The analysis of 15 continuous ECG recordings 20 min long, collected from healthy subjects, has shown that the QT-interval time series may be obtained automatically. The feasibility of automatic beat-by-beat measurement of the QT interval during sinus rhythm has been demonstrated. In simulated ECG with SNR=25, Q and T fiducial points were estimated with a precision of 0.5 ms and 2 ms, respectively. The introduction of the ACD (adapted cross difference) algorithm has doubled the precision of the R-WOT (R-wave occurrence time) estimator without augmenting the computation time. Preliminary results on the spectral content of QT variability indicate the presence of three major frequency bands (VLF: 0.001-0.05 Hz; LF: 0.06-0.12 Hz; and HF: 0.25-0.4 Hz), already described for the RR variability; the power distribution among these bands seems to be different for the two series.<
>