ISSN:
1573-8787
Keywords:
approximate entropy
;
central tendency measure
;
dynamical systems
;
heart rate variability
;
lyapunov exponents
;
nonlinear measures
;
nonlinear prediction
;
recurrence plot
;
stationarity
Source:
Springer Online Journal Archives 1860-2000
Topics:
Computer Science
Notes:
Abstract In this paper we investigate the requirementsfor the size of data set to be analyzed by a set of nonlinearmeasures. Time series from some standard nonlinear systems possessingchaotic behavior, as well as sinusoidal and random signals areconsidered. For Hénon, Kaplan-Yorke and logistic mapswe found the measures of correlation dimension (CD), approximateentropy (AE), Lyapunov exponents (LE), and deterministic ratio(DR) to be reliable for data length as short as 500 samples.On short-term heart rate variability (HRV) signal, AE and DRmeasures were able to distinguish between various experimentallymodified states of autonomic nervous system (ANS) controllingthe heart rate (HR). Thus, titling and parasympathetic blockademake the system more deterministic and reduce its entropy whilstsympathetic blockade makes it less deterministic.
Type of Medium:
Electronic Resource
URL:
http://dx.doi.org/10.1023/A:1026507317626
Permalink
