Tecnologia assistiva para pessoas com limitação motora severa usando processamento de potenciais de ação de unidades motoras de músculos faciais

• Ano de defesa: 2013
• Autor(a) principal: Pinheiro Júnior, Carlos Galvão
• Tipo de documento: Tese
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Uberlândia; BR; Programa de Pós-graduação em Engenharia Elétrica; Engenharias; UFU
• Palavras-chave em Português: Face - músculos; Interação homem-máquina; Eletromiografia; Tecnologia assistiva; Interação homem-computador; Eletromiografia facial; Electromyography; Assistive technology; Human-computer interaction; Facial electromyography; CNPQ::ENGENHARIAS::ENGENHARIA ELETRICA
• Link de acesso: https://repositorio.ufu.br/handle/123456789/14336

In some circumstances, a person may be deprived of natural abilities, such as walking and speaking, perhaps due to limb amputation, spinal cord injuries (SCI), or degenerative diseases. Assistive technology devices allows motor-impaired people to overcome their limitations promoting greater independence. Particularly suitable in the case of people with severe motor impairment, electrical biosignals have been successfully utilised to operate alternative communication devices. For over half a century, information extracted from the electromyographic signal for the purpose of operating a given device has not considered the information provided by the basic unit of the muscle: the motor unit. The objective of this study is to investigate how accessing information at motor unit level would improve the operator's performance during a given task. The hypothesis is that the proposed methodology would allow generating more precise control commands, when compared to traditional approaches relying on global information obtained by conventional electromyographic signal acquisition and processing. A system to detect motor unit action potentials from the electromyographic signal was devised, including the electrode design, and the performance evaluated by measuring the time taken to perform several cursor control tasks. The specications of the cursor control task were extracted from a dierent study, which used the traditional electromyographic signal-processing approach. Comparing the results from both studies proved that the novel approach provides better control than the traditional one, being 27% faster in the most dificult task.