A new model-based ectopic beat correction algorithm for heart rate variability

We propose a new ectopic beat correction algorithm for Heart Rate Variability (HRV) that is based on the Integral Pulse Frequency Modulation (IPFM) model. The model is used to infer the modulation signal that has generated the beats surrounding the ectopic beat. The ectopic beat is then removed and replaced with a normal beat at the time which least disturbs the rhythm identified with the IPFM model. We test the performance of the technique by removing simulated ectopy from real heart rate data from human subjects using a number of previously published correction techniques as well as our own. We show that our new technique performs significantly better than existing: techniques for low and high levels of ectopy.