Non-linear EEG analysis in children with attention-deficit/ hyperactivity disorder during the rest condition

Attention Deficit/Hyperactivity Disorder is a neuropsychiatric condition characterized by varying levels of hyperactivity, inattention and impulsivity. Electroencephalographic studies play an important role in brain-based cognitive analysis that can support clinical decisions and improve sensitivity and specificity of the decisions. This study investigates the non-linear dynamics of EEG signals regarding AD/HD and normal participants. Largest Lyapunov Exponent (LLE) and Approximate Entropy (ApEn) quantifies the non-linear chaotic dynamics of the signal. By assessment of ANOVAs test, it has been proven that the value of LLE in temporo-frontal cortices of brain represent significant difference between AD/HD and age- matched control group. Furthermore, the mean ApEn is significantly lower in AD/HD subjects. Evaluating of the features is performed by Probabilistic Neural Network. The results show that using nonlinear features along with probabilistic neural networks yielded a high accuracy of 87.5% for identification of ADHD in the nonlinear feature space discovered in this research.