On the presence of deterministic chaos in HRV signals

The presence of deterministic chaos in the instantaneous heart rate variability (HRV) signal is investigated by the reconstruction of a multi-dimensional signal using a time-delay embedding method. Four methods (autocorrelation, mutual information, average displacement and visual attractor inspection) are used to determine the time-delay parameter. The dimensionality of the embedding is tested using the nearest-neighbour and correlation-dimension methods, including a surrogate data test for the correlation dimension. An analysis is also performed for calculating the largest Lyapunov exponents. The results obtained in the time-delay embedding process support the conclusion that the HRV series could be chaotic.