Emotional Faces EEG Paradigm

Overview

A Brain Computer Interface is a device that uses the brain activity of a person as an input to select desired outputs on a computer. Using the brain activity of a person to identify his or her intentions gives the possibility of new communication pathways that can result in an increased quality of life for those patients with any degree of impairment on their sensory-motor abilities.

This project examines the current state of the art of non-invasive BCI, in particular those based on electroencephalogram (EEG) recordings, putting forwards novel methods applied to the BCI systems with the intentions to improve their performance and extend their use to any type of needs. Firstly, a novel paradigm based on emotions was proposed and evaluated with the intention of induce a stronger response on the user. Facial expressions give high volume of information and the human face recognition system involves unique properties that are not present in the recognition of other type of visual stimuli. These particular features may face stimuli suitable for BCI applications.

Another mechanism to enhance the efficiency and performance of BCI systems is by introducing novel signal processing methods. The main aim of this work is to develop an innovative and effective method to characterise functional connectivity on the brain. In order to understand how the different areas of the brain interact between them during the execution of a specific task the phase synchronization pattern from the EEG signals has been studied. From the analysis of the phase synchronization variation along time it can be observed the existence of quasi-stable states. The switching patterns between states are posteriorly translated into complex network graph metrics. The result of this transformation are used as features to feed different classification algorithms to distinguish between different tasks.