Fractal behaviour of pathological heart rate variability dynamics

Abstract

Heart rate variability analysis (HRV) is a well recognized tool in the autonomic control assessment. It has been suggested that nonlinear analysis of HRV might provide more valuable information than traditional linear methods. Several non linear fractal techniques recently gained wide interest: that based on indirect fractal dimension (FD) estimation from the 1/f spectral power relationship, and that based on a direct FD estimation from HRV time sequences. Aim of the study was to assess whether FD discriminates pathological HRV dynamics, comparing results with normal subjects and traditional linear indexes. We studied 7 groups of 10 ECG 24h-Holter recordings in normal and different pathologies: obstructive pulmonary disease, stroke, hypertension, post myocardial infarction, heart failure, heart transplanted. HRV was assessed by spectral power in very low, low and high frequency bands and standard deviation between normal beats. FD was estimated directly from the HRV sequences by Higuchi method (HM) and from the 1/f slope of spectral power relationship (beta). Results showed differences in the autonomic control impairments better described by FD than by traditional linear methods. Although HM and beta tried to measure the same FD property, the latter seemed to be rather insensitive to changes in autonomic control. These preliminary results clearly suggest that FD, estimated by HM, contains relevant information related to different HRV pathological dynamics.

Fractal behavior of heart rate variability during ECG stress test in cardiac patients

“…The fractal analysis was performed according to Higuchi's algorithm [7][8][9]. Let x(1),x (2),…,x(N) be the time series under investigation.…”

Relationships between linear and nonlinear indexes of heart rate variability in obstructive sleep apnea syndrome

“…In this context, several nonlinear methods, which have already shown to be useful in describing complex temporal processes, may be valuable to better describe HRV characteristics [4][5][6][7][8]. In particular, this work focuses on the use of the power-law exponent (β), the fractal dimension (FD) and the approximate entropy (ApEn) and draws attention to the existing relationships among them.…”