Dynamical measures for characterization of EEG registers in patients with Attention Deficit Hyperactivity Disorder treated with neurofeedback

Abstract

This article presents the results of the application of different measures of complexity based on nonlinear dynamics techniques, to evaluate the effect of a neurofeedback therapy in patients with Attention Deficit Hyperactivity Disorder (ADHD) utilizing electroencephalographic (EEG) registers as unique source of information. Every EEG register analyzed in this study contains 26 channels and was acquired in closed- and opened-eyes conditions, during pre- and post-treatment states. Four measures of complexity were applied: largest Lyapunov exponent, correlation dimension, Lempel-Ziv complexity and Hurst exponent. The purpose was to determine if these measures detect quantitative changes from the information contained in the EEG registers as consequence of the neurofeedback therapy. The results of this work indicate that Lempel-Ziv complexity and largest Lyapunov exponent could have a potential utility to detect quantitative differences between pre- and post-treatment in some specific channels. In contrast, correlation dimension and Hurst exponent provided scarce information about these differences. The most noticeable results were found in opened-eyes condition.