Discriminador binario de imaginación visual a partir de señales EEG basado en redes neuronales convolucionales

[EN] A Brain-Computer Intarface (BCI) is a technology that allows direct communication between the brain and the outside world without the need to use the peripheral nervous system. Most BCI systems focus on the use of motor imagination, evoked potentials, or slow cortical rhythms. In this work, the...

Descripción completa

Detalles Bibliográficos
Autores: Llorella, Fabio Ricardo, Iáñez, Eduardo, Azorín, José Maria, Patow, Gustavo
Tipo de recurso: artículo
Fecha de publicación:2021
País:España
Institución:Universitat Politècnica de València (UPV)
Repositorio:RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Idioma:español
OAI Identifier:oai:riunet.upv.es:10251/178698
Acceso en línea:https://riunet.upv.es/handle/10251/178698
Access Level:acceso abierto
Palabra clave:Brain-switch
Visual imagery
Convolutional neuronal network
Power spectral density
EEG
Discriminador binario
Interfaz cerebro-máquina
Red neuronal convolucional
Densidad potencia espectral
Descripción
Sumario:[EN] A Brain-Computer Intarface (BCI) is a technology that allows direct communication between the brain and the outside world without the need to use the peripheral nervous system. Most BCI systems focus on the use of motor imagination, evoked potentials, or slow cortical rhythms. In this work, the possibility of using visual imagination to construct a binary discriminator has been studied. EEG signals from seven people have been recorded while imagining seven geometric figures. Using convolutional neural networks it has been possible to distinguish between the imagination of a geometric figure and relaxation with an average success rate of 91 % with a Cohen kappa value of 0.77 and a percentage of false positives of 9 %.