Correction of erroneous and ectopic beats using a point process adaptive algorithm.

We present a new R-R interval correction procedure based on a point process model of the human heart beat. The algorithm combines an adaptive point process filter with a set of conditions on the probability of having a beat according to the model. This framework allows for correction of ectopic and erroneously detected beats in an on-line fashion, simultaneous with computation of instantaneous estimates of heart rate and heart rate variability. Results demonstrate the efficacy of the method, and show new heart rate and heart rate variability dynamics corrected for artifacts introduced by incorrect and/or irregular R-R intervals.