NONLINEAR-ANALYSIS OF EEG SLEEP STATES

Recent advances in the field of nonlinear dynamics have provided new conceptual models as well as novel analytical techniques applicable to neuropsychopharmacologic studies. One measurement technique that has been recently developed in an attempt to characterize nonlinear systems in physics and biology is the estimation of dimension. Dimension may be seen as a measure of the information required to describe the current behavior of a system. We have applied these techniques to the analysis of the sleep EEG, and have found that the dimension of rapid eye movement (REM) sleep is significantly higher than non-rapid-eye-movement (NREM) sleep. These data support a preliminary hypothesis that EEG dimension may represent the number of nonlinear modes activated in the brain. Thus, sleep states of low arousal or low input would be envisioned as having low dimension (e.g., slow-wave sleep) whereas increased arousal (REM) would activate more nonlinear modes. Although more investigations will be needed to explore this hypothesis, these studies suggest that further development of nonlinear approaches to the analysis of brain systems are likely to generate new clinical measures as well as new ways of viewing brain electrical function.