| Name: | Description: | Size: | Format: | |
|---|---|---|---|---|
| 15.49 MB | Adobe PDF |
Authors
Advisor(s)
Abstract(s)
Nas últimas décadas, diversas investigações têm realçado a importância de monitorizar a fadiga
muscular nos programas de reabilitação física, especialmente na recuperação de funções do
membro superior. No entanto, as terapias convencionais têm dificuldade em monitorizar a
fadiga muscular de forma precisa e em tempo real. Para colmatar esta dificuldade, a presente
dissertação explora uma alternativa inovadora, propondo o uso da análise das expressões
faciais como uma técnica promissora para a deteção da fadiga muscular. A investigação foca-se
em compreender quais são as variações nas expressões faciais que podem indicar fadiga
muscular durante a realização do exercício de flexão de bíceps. Este exercício é utilizado devido
à sua relevância clínica na reabilitação do membro superior, uma vez que simula movimentos
funcionais do dia a dia. A eletromiografia é empregue como uma ferramenta de validação,
servindo para confirmar a presença de fadiga muscular enquanto as expressões faciais são
monitorizadas. Os resultados indicam que apesar de não existirem diferenças estatisticamente
significativas entre séries na análise ANOVA, a análise das expressões faciais mais granular
permite concluir que as combinações de pontos faciais Boca aberta (0-17) e Franzir as
sobrancelhas (107-8) são indicadores relevantes na deteção da fadiga muscular. A compreensão
das expressões faciais como indicadores de fadiga não só permite melhorar os resultados
clínicos, mas também aumenta a adesão dos pacientes às terapias de reabilitação,
proporcionando uma experiência mais confortável e recetiva durante o processo de
recuperação. Assim, ao centrar-se neste método, a dissertação contribui para o
desenvolvimento de um sistema integrado que combina eletromiografia e reconhecimento
facial, com o objetivo de otimizar os processos de reabilitação e personalizar os programas
terapêuticos de acordo com as necessidades individuais dos pacientes, potencializando, dessa
forma, a sua recuperação. Este sistema visa ser integrado num cenário de reabilitação assistida
por robô, possibilitando que o braço mecânico se ajuste e adapte os movimentos e a
intensidade do tratamento com base nas informações obtidas pela avaliação da fadiga muscular
em tempo real.
In recent decades, numerous investigations have highlighted the importance of monitoring muscle fatigue in physical rehabilitation programs, particularly in the recovery of upper limb function. However, conventional therapies struggle to monitor muscle fatigue accurately and in real time. To address this challenge, the present dissertation explores an innovative alternative by proposing the use of facial expression analysis as a promising technique for detecting muscle fatigue. The research focuses on understanding which variations in facial expressions may indicate muscle fatigue during bicep curl exercises. This exercise is used due to its clinical relevance in upper limb rehabilitation, as it simulates functional movements performed in daily activities. Electromyography is employed as a validation tool, serving to confirm the presence of muscle fatigue while facial expressions are monitored. The results indicate that, although no statistically significant differences between series were found in the ANOVA analysis, a more granular analysis of facial expressions reveals that specific facial point combinations—such as Open Mouth (0-17) and Frowning (107-8)—are relevant indicators in detecting muscle fatigue. Understanding facial expressions as indicators of fatigue not only improves clinical outcomes but also enhances patient adherence to rehabilitation therapies by providing a more comfortable and engaging recovery experience. By focusing on this method, the dissertation contributes to the development of an integrated system that combines electromyography and facial recognition with the aim of optimizing rehabilitation processes and tailoring therapeutic programs according to the individual needs of patients, thereby enhancing their recovery. This system is intended to be integrated into a robot-assisted rehabilitation scenario, enabling the mechanical arm to adjust movements and treatment intensity based on real-time assessments of muscle fatigue.
In recent decades, numerous investigations have highlighted the importance of monitoring muscle fatigue in physical rehabilitation programs, particularly in the recovery of upper limb function. However, conventional therapies struggle to monitor muscle fatigue accurately and in real time. To address this challenge, the present dissertation explores an innovative alternative by proposing the use of facial expression analysis as a promising technique for detecting muscle fatigue. The research focuses on understanding which variations in facial expressions may indicate muscle fatigue during bicep curl exercises. This exercise is used due to its clinical relevance in upper limb rehabilitation, as it simulates functional movements performed in daily activities. Electromyography is employed as a validation tool, serving to confirm the presence of muscle fatigue while facial expressions are monitored. The results indicate that, although no statistically significant differences between series were found in the ANOVA analysis, a more granular analysis of facial expressions reveals that specific facial point combinations—such as Open Mouth (0-17) and Frowning (107-8)—are relevant indicators in detecting muscle fatigue. Understanding facial expressions as indicators of fatigue not only improves clinical outcomes but also enhances patient adherence to rehabilitation therapies by providing a more comfortable and engaging recovery experience. By focusing on this method, the dissertation contributes to the development of an integrated system that combines electromyography and facial recognition with the aim of optimizing rehabilitation processes and tailoring therapeutic programs according to the individual needs of patients, thereby enhancing their recovery. This system is intended to be integrated into a robot-assisted rehabilitation scenario, enabling the mechanical arm to adjust movements and treatment intensity based on real-time assessments of muscle fatigue.
Description
Keywords
Muscle fatigue Upper limb rehabilitation Electromyography Facial expression recognition Fadiga muscular Reabilitação do membro superior Eletromiografia Reconhecimento da expressão facial