Attractor reconstruction for plethysmographic biosignals

Analysis of biosignals is still an open problem, traditionally it has been tackled using frequency domain techniques. In this paper we propose to analyze the plethysmographic signal through its dynamical properties. Time series measured from healthy subjects were considered for proposing an analysis framework. We use delayed coordinates technique in order to reconstruct the state space dynamics of the attractor, then correlation dimension, Hurst and Lyapunov exponents were computed in order to further classify signals for different health conditions on patients.