Una prueba fuera del laboratorio de una interfaz cerebro computador basada en potenciales de estado estable evocados visualmente
Fecha
Autor corporativo
Título de la revista
ISSN de la revista
Título del volumen
Editor
Universidad Distrital Francisco José de Caldas. Colombia
Compartir
Director
Altmetric
Resumen
Descripción
Contexto: Los potenciales de estado estable evocados visualmente (SSVEP) surgen como una de las señales medidas del cerebro más promisorias para la implementación de interfaces cerebro computador (BCI), sin embargo las aplicaciones realizadas generalmente son limitadas a ambientes restringidos de laboratorio presentando mínimas pruebas en ambientes donde las condiciones son exigentes.Método: Se implementó un sistema BCI basado en SSVEP cuyas pruebas se realizaron en un ambiente no controlado con altos niveles de distracción y con la presencia de gran cantidad de público. Al usuario se le propone un laberinto donde un avatar debe ser llevado a un objetivo representado por una casa. Para mover el avatar el voluntario debe fijar su mirada en uno de los cuatro estímulos visuales, los estímulos visuales representan las cuatro direcciones: izquierda, arriba, derecha y abajo. El sistema fue usado sin ningún tipo de calibración.Resultados: 32 voluntarios usaron el sistema y 20 lograron el objetivo con una precisión mayor al 60%, incluyendo 9 con una precisión del 100%, 7 lograron el objetivo con una precisión menor al 60% y 5 se retiraron antes de conseguir el objetivo. Para los voluntarios que lograron una precisión mayor al 60% se alcanzó en promedio un tiempo de detección de 6,4s, una precisión de 79% y una razón de transferencia de información (ITR) de 8,78bits/s.Conclusiones: Se pudo mostrar un sistema BCI basado en SSVEP de bajo costo, probado en un evento público, sin procedimientos de calibración, de fácil instalación y usado por personas en un amplio rango de edades. Los resultados muestran que es factible llevar este tipo de sistemas a ambientes fuera del laboratorio.
Context: Steady State Visually Evoked Potentials (SSVEP) are brain signals which are one of the most promising signals for Brain Computer Interfaces (BCIs) implementation, however, SSVEP based BCI generally are proven in a controlled environment and there are a few tests in demanding conditions.Method: We present a SSVEP based BCI system that was used outside the lab in a noisy environment with distractions, and with the presence of public. For the tests, we showed a maze in a laptop where the user could move an avatar looking for a target that is represented by a house. In order to move the avatar, the volunteer must stare at one of the four visual stimuli; the four visual stimuli represent the four directions: right, up, left, and down. The system is proven without any calibration procedure.Results: 32 volunteers utilized the system and 20 achieved the target with an accuracy above 60%, including 9 with an accuracy of 100%, 7 achieved the target with an accuracy below 60% and 5 left without achieving the goal. For the volunteers who reached accuracy above 60%, the results of the performance achieved an average of 6,4s for command detections, precision of 79% and information transfer rate (ITR) of 8,78 bits/s.Conclusions: We showed a SSVEP based BCI system with low cost, it was proved in a public event, it did not have calibration procedures, it was easy to install, and it was used for people in a wide age range. The results show that it is possible to bring this kind of systems to environments outside the laboratory.
Context: Steady State Visually Evoked Potentials (SSVEP) are brain signals which are one of the most promising signals for Brain Computer Interfaces (BCIs) implementation, however, SSVEP based BCI generally are proven in a controlled environment and there are a few tests in demanding conditions.Method: We present a SSVEP based BCI system that was used outside the lab in a noisy environment with distractions, and with the presence of public. For the tests, we showed a maze in a laptop where the user could move an avatar looking for a target that is represented by a house. In order to move the avatar, the volunteer must stare at one of the four visual stimuli; the four visual stimuli represent the four directions: right, up, left, and down. The system is proven without any calibration procedure.Results: 32 volunteers utilized the system and 20 achieved the target with an accuracy above 60%, including 9 with an accuracy of 100%, 7 achieved the target with an accuracy below 60% and 5 left without achieving the goal. For the volunteers who reached accuracy above 60%, the results of the performance achieved an average of 6,4s for command detections, precision of 79% and information transfer rate (ITR) of 8,78 bits/s.Conclusions: We showed a SSVEP based BCI system with low cost, it was proved in a public event, it did not have calibration procedures, it was easy to install, and it was used for people in a wide age range. The results show that it is possible to bring this kind of systems to environments outside the laboratory.
Palabras clave
BCI, Brain Computer Interface, MEC, Minimum Energy Combination, SSVEP, Steady State Visually Evoked Potentials, BCI, Combinación de Mínima Energía, Interfaz cerebro computador, MEC, potenciales de estado estable evocados visualmente, SSVEP.