Local Holder regularity-based modeling of RR intervals
Abstract
We analyze the local regularity of RR traces from ECG through the computation of the so-called Hölder exponents. These exponents are at the basis of multifractal analysis, which has been shown to be relevant in the study of RR data. While multifractal analysis yields a global picture of the (statistical) distribution of regularity, we focus here on its time evolution. We show that this evolution is strongly correlated with the signal itself, a feature that seems to have remained unnoticed so far. We use this fact to build realistic synthetic RR traces.
Domains
Probability [math.PR]
Fichier principal
Local-Holder-regularity-based-modeling-of-rr-intervals.pdf (1.61 Mo)
Télécharger le fichier
Origin : Files produced by the author(s)