Complexity analysis of electrocardiographic signals.

Two types of electrocardiographic data series were investigated using appropriate tests based on a selection of semi-quantitative analysis algorithms. Distribution histograms, power spectra, auto-correlation functions, state-space portraits, Lyapunov exponents and wavelet transformations were applied to electrocardiograms of normal and stressed subjects. Statistical analysis using the Student's t-test revealed significant and non-significant alterations in stress-loaded cases compared to normal ones. Higher levels of adrenaline may account for a more complex dynamics (deterministic chaos) revealed in the stressed subjects.