Design and applications of brain-computer interfaces


Sebastian Halder:
 
Tutorial Title:

Design and applications of brain-computer interfaces

Abstract:
Brain-computer interfaces (BCI) provide a direct method for processing neurophysiological signals, such as the electroencephalogram (EEG), for restoring, repairing or augmenting motor or cognitive functions.  Successful implementation of a BCI involves choosing the appropriate control signal, feature extraction, machine learning method and interface design. Each component should be chosen in accordance with the intended user group and application. This is particularly challenging when developing a BCI for persons with severe motor impairments.

We will discuss the components of a BCI as well as the potential applications and limitations. In the practical part of the tutorial, we will set up a BCI based on the visual P300 event-related potential. This  will provide insights on (1) the challenges of preparing an EEG recording and ensuring that the brain signals are recorded with high  quality, (2) analyzing calibration data using Matlab scripts to train a classifier that can detect the user’s control signal (3) and configuring the BCI system for online control of a spelling system or a device such as a robot.