Algorithmic testing for QT interval measurement

The PTB diagnostic ECG database has been used for testing the fully automated program for the identification and measurement of QT interval. The database consists of 549 ECG recordings of 294 subjects. The testing algorithm uses the Padova program, which identifies all QRS complexes analyzing a global spatial velocity and the ECG signal. Different strategies for the choice of the optimal QT interval have been compared, and the optimization of T end in the single leads was performed using the well annotated CSE database. This paper indicates that the more reliable and accurate estimates of the QT interval is the median value of all the measurements from the analysis of the 12 ECG leads (QT global). In addition, the Challenge and the public PTB database reveal their potentiality for the improvement of ECG classifiers and in particular of QT estimators. However, for clinical evaluation it is necessary to have a more consistent "gold standard".