Evaluation of nonlinear dynamics in postural steadiness time series

Fractal and correlation dimensions have been computed for time series obtained from tests of balance (postural steadiness). Although these measures appear to be reliable and differentiate subject groups, it has become clear that random (noise) time series may have finite dimensions and appear to demonstrate dynamics characteristic of nonlinear systems. Consequently, it is necessary to apply a test to distinguish a time series with putative nonlinear dynamics from random noise. A simple predictor was utilized to compare center of pressure (COP) time series with surrogate data constructed to have similar time and frequency domain characteristics. It was found that the original time series was more predictable than the surrogate data, suggesting that the COP data is derived from a nonlinear system.