EEG analysis in patients with schizophrenia based on Lyapunov exponents

Abstract

The present work focused on study of the chaotic nature of the electroencephalogram (EEG) signal in schizophrenia by assessing Lyapunov exponents, which can serve as indicators of specific brain function suffered from the disease. EEG series were collected from 45 patients with schizophrenia (aged between 10 and 14 years) and age-matched 39 healthy adolescents. EEG recordings were acquired using 16 channels. To calculate the Largest Lyapunov exponent, the Kantz, Rosenstein and Wolf algorithms were utilized, and the spectrum of exponents was calculated with the Sano-Sawada algorithm. The results reveal that schizophrenia is characterized by the Largest Lyapunov Exponents estimated via the Rosenstein algorithm. The Lyapunov spectrum provides an opportunity to increase the classification accuracy due to a larger number of significant channels. Thus, non-linear EEG analysis, based on Lyapunov exponents, is considered useful for neurodynamic research of patients with schizophrenia.