Abstract
The behaviour of many natural complex systems is characterized by
nonstationarities and phase shifts making a conventional analysis of
periodicities not fully reliable. Recently, the method of Phase
Rectified Signal Averaging (PRSA) 1 has been introduced for the
extraction of non-stationary oscillations out of noisy signals with
varying mean. As an example, PRSA was shown to be superior in risk
classification of sudden cardiac death after initial myocardial
infarction 2. The main advantage of the PRSA is its capability to
analyze separately periodicities occurring around increases (or,
alternatively, decreases) of the signal. We now suggest a multivariate
form of PRSA to study the relationships (i.e., interactions, partial
syncronization, etc.) between two or more complex signals. We compare
this method with cross correlation and cross spectra techniques and also
discuss the application of multivariate PRSA in a recent baroreflex
regulation study.
1) A. Bauer et al., Physica A 364, 423 (2006)\\
2) A. Bauer et al., The Lancet 367, 1674 (2006)
Users
Please
log in to take part in the discussion (add own reviews or comments).