Identificação de Estado Mental de Atenção Através do EEG para Aplicação em Treinamento Neurofeedback

• Ano de defesa: 2019
• Autor(a) principal: Casagrande, Wagner Dias
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal do Espírito Santo; BR; Mestrado em Engenharia Elétrica; Centro Tecnológico; UFES; Programa de Pós-Graduação em Engenharia Elétrica
• Palavras-chave em Português: Discrete Wavelet Transform (DWT); Electroencephalography; Supporting Vector Machine (SVM); Attention Deficit Hyperactivity Disorder (ADHD); Neurofeedback; Transformada wavelet discreta (DWT); Máquina de vetores de suporte (SVM); Transtorno de déficit de atenção ehiperatividade (TDAH); Eletroencefalografia; Fourier, Transformações de; Aprendizado do computador; Distúrbio da falta de atenção com hiperatividade; Sistemas de controle biológico; Eletrônica Industrial, Sistemas e Controles Eletrônicos; 621.3
• Link de acesso: http://repositorio.ufes.br/handle/10/11166

Attention Deficit Hyperactivity Disorder (ADHD) is a disorder associated with neurobiolog-ical and genetic factors, beginning in childhood and can persist in adulthood, compromisingthe functioning of the person in various sectors of his life. The worldwide prevalence ratesfor ADHD are around 5.3% in children and adolescents and 2.5% in adults. Studies suggestthat 40 to 60% of affected children continue to have the disorder in adulthood. Althoughpharmacological treatment has proved effective, it still has disadvantages, such as sideeffects (anorexia), abdominal pain, headache and insomnia. Given the relatively high rateof residual symptoms, the disability generated by this disorder and the possible resistanceto pharmacological treatment, it is necessary to combine the available therapeutic arsenalwith new non-pharmacological methods. Neurofeedback has been shown through numerousstudies to significantly improve attention and provide equivalent improvements to stimulantmedications for attention deficit hyperactivity disorder (ADHD) children. This work aimsto develop a neurofeedback system aimed at supporting the conventional treatment ofchildren with ADHD, as well as the development of a serious game in order to generatevisual feedback. The same uses the electroencephalographic signals captured throughthe EEG to classify the mental task of attention. For the analysis and processing of thesignals, the Welch method and the discrete wavelet transform (DWT) were used for theextraction of characteristics, the linear discriminant analysis (LDA) was applied to reducethe dimensionality of the characteristics for each electrode and the Machine of SupportVectors (SVM) was used to classify the attention state. In this way, conventional drug-basedtreatment can be supplemented to improve patients’ final response. An experimental testprotocol was proposed and the results demonstrated that the system was able to acquire,process and classify the signals. Analysis of the accuracy of the used classifier showedthat the system was able to identify the instants in which the subjects were in a state ofattention and relaxation, with the best accuracy being 90% for the Welch method. Theresults of the processing steps of the biological signals, as well as the accuracy valueswere shown in accordance with the studied literature. Future work involves improving thetechniques used in signal processing, developing a complex environment generated by thevirtual reality technique and validating the system with children who have ADHD.