We investigated the threshold dependent complexity of a physiological signal, such as heart rate
variability (HRV). Complexity of a binary symbolic sequence reconstructed from HRV of length 2048
was estimated for a normal group with stable sinus rhyth...
We investigated the threshold dependent complexity of a physiological signal, such as heart rate
variability (HRV). Complexity of a binary symbolic sequence reconstructed from HRV of length 2048
was estimated for a normal group with stable sinus rhythm and an abnormal group with congestive
heart failure or atrial fibrillation by using complexity measure, which defines the interrelation between
the uncertainty and disequilibrium of information on HRV. We found that the complexity of
the normal was higher than that of the abnormal in a range of lower thresholds (10 20 msec). In
the analysis of pattern distribution among groups, the distribution of the normal is more uniform
than that of the abnormal in a range of threshold larger than 25 msec, which confirmed that the
variation of heart rate of the normal is more complex than that of the abnormal.0