Dynamical Properties of nystagmus

Abstract

By the methods of the time series analysis we investigate human vestibular and optovestibular nystagmus time series. They can be stationary or nonstationary and are between the classes of purely periodic or purely chaotic time series. The amplitudes of the autocorrelation functions of the time series have an unexpected peak between 7 and 9 seconds on the time axis. The singular spectrum analysis shows that the important information about the time series is concentrated in the first four to six principal components. The influence of the sensory input modality, which can be changed by opening or closing eyes of the investigated persons, leads to changes in the histogram, but leaves almost unchanged other important characteristics of the time series such as the statistical dimension or the positions of the dominant frequencies in the power spectrum. Thus the behavior of the studied time series shows a presence of nonlinearities which could play an important role for realization of the nystagmus phenomenon.