Statistical Properties of the Interbeat Interval Cascade in Human Subjects

Abstract

Statistical properties of interbeat intervals cascade are evaluated by considering the joint probability distribution $P(\Delta x_2,\tau_2;\Delta x_1,\tau_1)$ for two interbeat increments $\Delta x_1$ and $\Delta x_2$ of different time scales $\tau_1$ and $\tau_2$. We present evidence that the conditional probability distribution $P(\Delta x_2,\tau_2|\Delta x_1,\tau_1)$ may obey a Chapman-Kolmogorov equation. The corresponding Kramers-Moyal (KM) coefficients are evaluated. It is shown that while the first and second KM coefficients, i.e., the drift and diffusion coefficients, take on well-defined and significant values, the higher-order coefficients in the KM expansion are very small. As a result, the joint probability distributions of the increments in the interbeat intervals obey a Fokker-Planck equation. The method provides a novel technique for distinguishing the two classes of subjects in terms of the drift and diffusion coefficients, which behave differently for two classes of the subjects, namely, healthy subjects and those with congestive heart failure.