Identification of ADHD Cognitive Pattern Disturbances Using EEG and Wavelets Analysis

Abstract

Recurrent investigations about human cognitive patterns evidenced by electrical brain signals have been performed with the use of non-invasive electroencephalographic instrumentation (EEG). In this context, this work proposes a methodology which is characterized by using EEG technique in the investigation of brain signal patterns in learners with Attention Deficit Hiperactivity Disorder - ADHD, evoked during the course of a typical activity called "Oddball Paradigm". Therefore, was applied the methodology that consisted in the selection of a group of children previously diagnosed with ADHD and a control group. During the work environments are also presented that have been implemented for the acquisition, preprocessing and visualization of brain electrical biossignals as well as environments for visual stimulation of Oddball Paradigm. With software support it was made using of mathematical modeling of Wavelet Transform to the decomposition of EEG signals. It was also applied the data classification method, using the Support Vector Machine (SVM) technique to establish patterns indicative of ADHD and normality based on the energy and power information extracted from the application of Morlet Wavelet Transform to the EEG records. It was obtained as a result of the applied methodology, the construction a model capable of classifying ADHD individuals and control group with 94.74% of accuracy.