The Exponential Distribution Applied to Nonequidistantly Sampled Cardiovascular Time Series

Abstract

Beat-to-beat fluctuations in cardiovascular time series comprise different frequency components which can be employed to describe autonomic regulatory processes. The Exponential Distribution (ED) is presented here as a specific time-frequency distribution which has the potential to describe the time-related changes in the frequency content of these cardiovascular fluctuations. The ED has as advantage that it gives a good suppression of the cross terms, a characteristic feature of bilinear time-frequency distributions. An implementation to apply the ED to nonequidistantly sampled cardiovascular time series is provided. Applications of the ED to various clinical and experimental human cardiovascular time series show that the ED can be an important aid to describe and interpret time-varying frequency components of cardiovascular signals such as heart rate, interbeat interval, blood pressure, and respiration.