Reliable real-time calculation of heart-rate complexity in critically ill patients using multiple noisy waveform sources
This paper presents a novel algorithm to detect onset and duration of QRS complexes. After low-pass filtering, the ECG signal is converted to a curve length signal by a transform in which a nonlinear scaling factor is introduced to enhance the QRS complex and to suppress unwanted noise. Adoptive thresholds are applied to the length signal to determine the onset and duration of the QRS complex. The algorithm was evaluated with the complete set of single channel ECGs (signal 0) from the MIT-BIH Arrhythmia Database, and achieved a gross QRS sensitivity of 99.65% and a gross QRS positive predictive accuracy of 99.77%. The QRS onsef determination is very stable and is insensitive to QRS morphology change. The noise tolerance of the algorithm was evaluated using the MIT-BIH Noise Stress Test Database. The C source code for the single-channel algorithm has been contributed to PhysioToolkit and is freely available from PhysioNet (www.physionet.org).