Conectividade cerebral para o controle online de interfaces cérebro-máquina

Detalhes bibliográficos
Ano de defesa: 2019
Autor(a) principal: Saito, Éric Kauati
Orientador(a): Não Informado pela instituição
Banca de defesa: Não Informado pela instituição
Tipo de documento: Dissertação
Tipo de acesso: Acesso aberto
Idioma: por
Instituição de defesa: Universidade Federal do Rio de Janeiro
Brasil
Instituto Alberto Luiz Coimbra de Pós-Graduação e Pesquisa de Engenharia
Programa de Pós-Graduação em Engenharia Biomédica
UFRJ
Programa de Pós-Graduação: Não Informado pela instituição
Departamento: Não Informado pela instituição
País: Não Informado pela instituição
Palavras-chave em Português:
EEG
Link de acesso: http://hdl.handle.net/11422/13223
Resumo: The classification of different Motor Imagery (MI) tasks of the same limb remains a challenge using electroencephalography (EEG), since its representation in the motor cortex area is very close spatially. Methods that consider the information flow, like Partial Directed Coherence (PDC), have shown to be promising for brain-computer interface (BCI) of the same limb. The objective of this thesis is to identify the MI of elbow flexion and extension to activate a robotic arm with a synchronous BCI. Parameters and preprocessing methods were study in offline classification before their online application. EEG signals of two volunteers were recorded performing during the MI of the elbow flexion and extension, following a robotic arm. For each volunteer, the connectivity function was used to obtain the characteristics for artificial neural network (ANN) training. Four online session were performed. In online validation (3 class: MI elbow flexion and extension and rest), the best accuracy for the first volunteer was 48.0% (third session), and 58.0% for the second volunteer (fourth session). Therefore, directed connectivity looks promising for the development of online BCI for the classification of MI of the same limb.