Por favor, use este identificador para citar o enlazar este ítem:
https://hdl.handle.net/11000/4817
Desarrollo de nuevos esquemas de entrenamiento y modalidades de retroalimentación para facilitar la
operación de interfaces cerebrocomputadora
Título : Desarrollo de nuevos esquemas de entrenamiento y modalidades de retroalimentación para facilitar la
operación de interfaces cerebrocomputadora |
Autor : Angulo Sherman, Irma Nayeli |
Tutor: Azorín Poveda, José María |
Departamento: Departamentos de la UMH::Ingeniería de Sistemas y Automática |
Fecha de publicación: 2018-07-12 |
URI : http://hdl.handle.net/11000/4817 |
Resumen :
Una interfaz cerebro-computadora (brain-computer interface o BCI) basada en
imaginación motora (IM) es un sistema que identifica dichos eventos imaginados
en las señales cerebrales y los traduce en comandos para un dispositivo externo.
Este sistema requiere que el usuario sea capaz de modular su ac... Ver más
A brain-computer interface or BCI that is based on motor imagery (MI) is a
system that identifies such events from brain signals and translates them into
commands for an external device. This system requires that the user is capable
of modulating brain activity at will. Hence, there is the need of a training phase
that incorporates strategies for enhancing the learning of such modulation. For
this reason, this thesis is oriented to the proposal and evaluation of new training
schemes for MI-based BCIs.
First, the accuracy of a training scheme in which functional electrical stimulation
(FES) is applied before the use of a BCI that provides visual feedback is
evaluated. A comparison is performed between this scheme and other training
modalities without FES, in which either visual or auditive feedback is given. As
result, FES shows some potential in enhancing BCI operability in some subjects.
Then, in order to understand the effect of the described training schemes, the
coherence from electroencephalographic signals (EEG) was studied. The results
from this analysis showed the correlation between the accuracy and the coherence
of the sensorimotor rhythm at the centroparietal and centrofrontal regions that
are contralateral to the MI. Hence, the coherence could be used as a metric to
evaluate BCI training schemes. Moreover, the usefulness of accuracy predictors
based on EEG data was assessed, revealing that such metrics do not seem to provide
enough information to infer if a user would be able to control a BCI during
training.
Furthermore, this thesis evaluated the enhancing effect of providing transcranial
direct current stimulation (tDCS) in BCI systems. For that, the effect of
supplying different current intensities over the classification of MI and EEG was
evaluated in two different tDCS stimulation montages: an array that stimulates
the motor cortex and another that influences the cortico-cerebellar pathway. In
particular, the cortico-cerebellar montage showed some potential for enhancing
the detection of right-hand MI. However, evaluating the delivery of higher current
intensities is still required for both montages. Finally, the EEG spectral chan-
ges that were produced during gait MI were analyzed with the aim of proposing
and evaluating EEG features for their classification in a BCI system.
|
Palabras clave/Materias: Rehabilitación médica Tratamiento de señales Neurociencia |
Área de conocimiento : CDU: Ciencias aplicadas: Ingeniería. Tecnología: Ingeniería mecánica en general. Tecnología nuclear. Electrotecnia. Maquinaria: Ingeniería eléctrica. Electrotecnia. Telecomunicaciones |
Tipo de documento : info:eu- repo/semantics/bachelorThesis |
Derechos de acceso: info:eu-repo/semantics/openAccess |
Aparece en las colecciones: Tesis doctorales - Ciencias e Ingenierías
|
La licencia se describe como: Atribución-NonComercial-NoDerivada 4.0 Internacional.