On the Variability of Chaos Indices in Sleep EEG Signals

TL;DR

This study investigates how chaos indices like Lyapunov exponent and correlation dimension vary across sleep stages and subjects, revealing potential markers for sleep disorders using EEG data.

Contribution

It provides a comprehensive analysis of the variability of multiple chaos indices in sleep EEG signals across different subjects and conditions.

Findings

Chaos indices vary significantly across sleep stages.

Indices can distinguish healthy subjects from those with sleep disorders.

Empirical histograms show clear differences in chaos measures.

Abstract

Previous researches revealed the chaotic and nonlinear nature of EEG signal. In this paper we inspected the variability of chaotic indices of the sleep EEG signal such as largest Lyapunov exponent, mutual information, correlation dimension and minimum embedding dimension among different subjects, conditions and sleep stages. Empirical histograms of these indices are obtained from sleep EEG of 31 subjects, showing that, with a good accuracy, these indices in each stage of sleep vary from healthy human subjects to subjects suspected to have sleep-disordered breathing.