Our study shows that some but not all computer programs for the interpretation of ECGs perform almost as well as cardiologists in identifying seven major cardiac disorders.

Introduction: The aim of this study was to determine whether impaired adaptation of the QT interval to changes in heart rate predicts sudden death after an acute myocardial infarction. Methods and Results: The Groupe d'Etude du Pronostic de l'Infarctus du Myocarde (GREPI) trial was a prospective multicenter study designed to evaluate the long‐term outcome of myocardial infarction. QT dynamicity was evaluated in 265 patients by analyzing 24‐hour Holter recordings obtained 9 to 14 days after myocardial infarction. The linear regression slope of QT intervals measured to the apex and to the end of the T wave (QTe) plotted against RR intervals was calculated using a dedicated Holter algorithm. The value of QT/RR in predicting sudden death and total mortality was compared with those of ejection fraction, heart rate variability, and late potentials. Mean follow‐up was 81 ± 27 months. There were 73 deaths, of which 23 were sudden. Of all the parameters, an increased diurnal QTe/RR slope (>0.18) was the strongest independent predictor of sudden death (relative risk 6.07, confidence interval 1.48–24.95, P = 0.01 ). Conclusion: Increased diurnal QTe dynamicity is independently predictive of sudden death among patients with myocardial infarction. This simple parameter may help to stratify risk and select patients who may benefit from antiarrhythmic prophylaxis. (J Cardiovasc Electrophysiol, Vol. 14, pp. 227‐233, March 2003)

Synthesis of the 12-lead ECG has been investigated in the past decade as a method to improve patient monitoring in situations where the acquisition of the 12-lead ECG is cumbersome and time consuming. This paper presents and assesses a novel approach for deriving 12-lead ECGs from a pseudoorthogonal three-lead subset via generic and patient-specific nonlinear reconstruction methods based on the use of artificial neural-networks (ANNs) committees. We train and test the ANN on a set of serial ECGs from 120 cardiac inpatients from the intensive care unit of the Cardiology Hospital of Lyon. We then assess the similarity between the synthesized ECGs and the original ECGs at the quantitative level in comparison with generic and patient-specific multiple-regression-based methods. The ANN achieved accurate reconstruction of the 12-lead ECGs of the study population using both generic and patient-specific ANN transforms, showing significant improvements over generic (p -value < or = 0.05) and patient-specific ( p-value < or = 0.01) multiple-linear-regression-based models. Consequently, our neural-network-based approach has proven to be sufficiently accurate to be deployed in home care as well as in ambulatory situations to synthesize a standard 12-lead ECG from a reduced lead-set ECG recording.