Accessing physiological complexity of HRV by using threshold-dependent symbolic entropy

There has been considerable interest in quantifying the complexity of physiological time series, such as heart-rate variability. By using threshold-dependent symbolic entropy, which is based on symbolic nonlinear time series analysis, it is possible to consistently quantify the complexity of physiological signals and noisy signals. We calculated the information entropy of Shannon and entropy rate of Kolmogorov-Sinai, with correction terms to compensate for the finiteness of symbolic sequences extracted from heartbeat. These corrected symbolic entropies consistently separate the normal and pathologic physiological symbolic sequences within a range of thresholds. We also discuss the problem of suitable generating partition, the effect of randomization, and the role of random noise in estimating the complexity of physiological symbolic sequences.