Quantification of the electrocardiographic ventricular repolarization involving the T-U wave complex is usually performed with reference to the axis of the T wave and the QT interval duration. A novel quantitative approach to improve the description of ventricular repolarization was applied to the digitized electrocardiograms of 423 normal subjects. Six electrocardiographic repolarization characteristics were identified: duration, rate, area, symmetry, late phenomena, and interlead heterogeneity. A computer algorithm was designed to automatically interpret the electrocardiographic repolarization segment and measure 11 variables that quantified these repolarization characteristics. The application of redundancy-reduction techniques selected a final set of seven variables that were used in the statistical analysis. The QT interval, which was included in the initial group of variables, was replaced by the time interval between S wave offset and T wave maximum. All selected electrocardiographic variables were independent of age (r2<0.11) and body surface area (r2<0.03); all except the early duration variable were heart rate-and QT interval-independent (r2<0.2, r2<0.13, respectively; and most were uncorrelated to each other. A comparison of repolarization characteristics by gender revealed that repolarization duration was significantly more prolonged (p<0.0001) in women than in men. This multidimensional quantitative approach conveys a new and more complete description of the repolarization process and provides an electrocardiographic repolarization database in normal subjects as a reference standard for identifying patients with disordered repolarization. (Circulation 1989;80:1301-1308 V nentricular repolarization is a complex electrical phenomenon that has been studied theoretically and experimentally.1-5 The T-U wave complex on the surface electrocardiogram (ECG) is the integrated signal of this repolarization process, and it has several distinct morphologic features. Conventionally, the ECG repolarization duration is quantified by measuring the QT interval or the heart rate-corrected form (QTc), both of which are used clinically to assess the propensity to ventricular arrhythmias in certain subsets of patients. accurate quantification of dilferent repolarization features has been possible.The purpose of our study is to identify and quantify several descriptive ECG features of ventricular repolarization using a validated computer algorithm for the analysis of digitized surface ECGs. This morphologic database of ECG ventricular repolarization parameters obtained in a large normal population can be a reference standard for identifying patients with disordered repolarization. Methods PopulationThe study population comprised respondents to local advertisements offering free physical examinations and 12-lead ECG recordings to healthy individuals. Individuals were excluded if they had one or more of the following: 1) medical history suggesting any organic heart disease, atrial flutter or fibrillation, sustained ventricu...

These findings indicate that 1) quantification of the dynamic relation between ventricular repolarization and RR cycle length can be obtained on a large number of Holter-recorded heart beats; 2) beta-blockers reduce the RTm/RR slope in normal patients; and 3) LQTS patients have an exaggerated delay in repolarization at long RR cycle lengths.

No abstract

The R-R interval measurement from digitized electrocardiograms (ECG) contains an error due to the finite sampling frequency which may jeopardize the beat-to-beat analysis of the heart rate. In this paper, we develop a model to describe and quantitate this error. The "measured" R-R interval is modeled as the sum of the "true" R-R interval and of the error of measurement. The first and second order statistics of the error are computed in order to investigate its influence on the heart rate variability (HRV) power spectrum. They are found to be only functions of the ECG sampling frequency and, in particular, the power spectrum of the error contributes an additive high-pass filter-like term (colored noise) to the power spectrum of the HRV. The accuracy of the model is tested via a simulation procedure. The model indicates that the relative balance between the HRV and the error power spectra is important and should be checked before any variability analysis on the heart rate. This balance may be favorable to the error when 1) the sampling frequency of the ECG is too low, and/or 2) the variability of the heart rate is too little. In these cases, the HRV spectrum analysis may not give reliable results. Two tests are proposed in order to evaluate the error influence either in specific frequency bands or in the total frequency range.

No abstract