Permutation Entropy Analysis of Heart Rate Variability for the Assessment of Cardiovascular Autonomic Neuropathy

Abstract

Heartbeat time series are clinically relevant biosignals as they hold substantial information regarding cardiovascular neuroregulatory mechanisms. Disorders of cardiovascular autonomic regulation occur during the pathogenesis of several diseases and are clinically important because of its life threatening consequences. Heart rate variability (HRV) analysis is widely used to characterize cardiovascular autonomic function (CAF), however its actual power relies on the selection of suitable features providing reliable information of the underlying neural modulation. Therefore, in this work we aim to explore the potential of permutation entropy (PE) analysis of HRV for evaluating neuropathological changes on CAF. For this purpose we quantified PE to assess the complexity of ordinal patterns from five minutes interbeat intervals series, in healthy subjects and patients with type 1 diabetes mellitus, including patients with cardiovascular autonomic neuropathy (CAN). We found that some PE indicators were significantly lower in the group of diabetic patients with CAN compared to those calculated for the control group, allowing differentiation between them. We hypothesize this happens due to a decrease of complexity in cardiac intervals series induced by physiological changes imposed by the disease. It was found that correlations of PE measures with standard HRV indicators depended only on the temporal scales considered to create the patterns, regardless their length. We concluded that PE analysis of HRV is an adequate and promising method for the assessment of CAN. Further research should be performed in order to unravel physiological meaning of this feature.