Automated adaptive heart interbeat time extraction from long term noisy and variable ECG signals.

We describe the development of an automated, adaptive method to obtain the time interval between successive heart beats from noisy and highly variable electrocardiography signals. These interbeat time series are critical to the fractal characterization of cardiac health. When the biophysical measurement is severely tainted with noise from multiple sources, there is a need for algorithms to robustly extract the important patterns from the signal in question. The proposed method yields interevent times that are in close agreement with those obtained by manual extraction, while significantly reducing the requisite processing time. The algorithm can be extended to other applications where simple threshold-based extraction methods fall short.