Classification of physiologic and synthetic heart rate variability series using base-scale entropy

The base-scale entropy method was used as a measure to classify physiologic and synthetic heart rate variability series. This method enables analyzing very short, non-stationary, and noisy data. We used it to analyze short-term heart rate variability series. The results show that our method can effectively detect the complex dissimilarity of physiologic time series in different physiologic/pathologic states. We then applied it to the CinC 2002 test datasets. Using the base-scale entropy, we correctly classified 43 of 46 (93%) time series. In combination with time domain analysis, we correctly classified all time series.