An algorithm for robust detection of QRS onset and offset in ECG signals

This paper proposes a robust algorithm for QRS onset and offset detection in ECG signals. It mainly consists of the computation of an indicator related to the area covered by the QRS complex envelope. This algorithm was initially developed for the detection of T wave end, whose robustness and efficiency for that purpose have been previously reported. Its performances for QRS onset and offset detection have been evaluated using the PhysioNet QT database. For QRS offset detection, this new algorithm improves the best results reported in the publications about other algorithms evaluated on the same database, and for QRS onset detection, its performance is close to the best published results.