Estimating the complexity of heart rate fluctuations - an approach based on compression entropy

Abstract

Heart rate exhibits spontaneous fluctuations that are mainly modulated by control loops within the autonomic nervous system. Assessing the dynamics of heart rate fluctuations can provide valuable information about regulatory processes and patho-physiological behavior. In this paper heart rate fluctuations and its entropy are assessed using an algorithmic information theoretic concept applying a data compressor. First, the beat-to-beat fluctuations of heart rate are binary coded for decreases and increases, respectively. Subsequently, those symbol sequences are compressed using the LZ77 algorithm. The ratio of the length of the compressed sequences to the original length is used as an estimate of entropy. We investigated the compressibility of heart rate fluctuations in athletes before, during, and after a training camp. Heart rate time series were obtained from ECGs recorded over 30 minutes under supine resting conditions. We found a significant entropy reduction during the training camp, reflecting the effects of physical fatigue. In conclusion, the compression entropy seems to be a suitable approach to assess the complexity of heart rate fluctuations.