Análise de sinais se EEG utilizando a transformada Wavelet como subsídio para identificação de TDAH em crianças de idade escolar

Abstract

Attention Deficit/Hyperactivity Disorder (ADHD) is a neurobiological disorder that is characterized by a persistent pattern of inattention/hyperactivity-impulsivity. In school-age children, the influence of this disorder may result in low academic performance, however, the main factor is the interference in the individual's social, academic and professional life. Thus, this work proposes to develop an analysis system based on the Electroencephalogram (EEG) signal to encourage the development of tools to identify suggestive signs related to ADHD in school-age children. For this, the classifier is based on the Threshold technique using Discrete Redundant Wavelet Transform to extract signal features. The simulation environment used was MATLAB software (2015a). The analyzed dataset was from the IEEE Dataport database. To achieve the objective of the work, the delta and theta frequency ranges of the wavelet coefficients were used as parameters for the threshold method, in addition, the analyzed electrodes were from the frontal region of the brain. The proposed model performed with a sensitivity of 88.58 % and positive predictability of 73.26 % for a set of 40 analyzed data. Among the identified aspects, it appears that the performance of the algorithm was satisfactory, however, for a small volume of data.