The Neurophysiological Biomarker Toolbox (NBT)



Brain computer interface (BCI): Introduction

Owing to the developments in the neurosciences and computer technology, it has become possible to link brain activity to the operation of computers and devices, creating a direct communication channel between mind and machine (for reviews, see Van Gerven et al., 2009 and Wolpaw et al. 2002). This is possible with a brain-computer interface (BCI), which exploits the fact that certain aspects of brain activity are linked to specific mental states and processes, called 'signatures' (Van Gerven, 2009). A BCI is a combination of techniques for recording brain activity, extracting and processing signatures, and translating aspects of the signature into computer commands, which are fed back to the user (Fig. 1). There are BCIs that partially restore movement and/or communicative capabilities in paralysed patients (Birbaumer & Cohen 2007), as well as BCIs that explore new ways of playing computer games (Nijholt et al. 2009).

This wiki is about a BCI introduced in 1988 (Farwell & Donchin), which allows a subject to write letters on a screen. It is based on electroencephalography (EEG).

How can brain activity be linked to computer commands?

The operation of a BCI generally involves several steps (see Figure 1):

  1. measurement of brain activity
  2. signature extraction
  3. translation of (aspects of) the signature into computer commands
  4. feedback


Figure 1: An example of a brain-computer interface cycle

Many techniques for measuring brain signals can be used in a BCI (fMRI, NIRS, EEG, MEG, etc.) and the pre-processing of these measurement data can differ. However, the primary task of any BCI is to correctly identify the signature and base computer commands on this that satisfy the BCI user's desire. In the case of the P300 speller, a characteristic attention-related brain signal known as the P300 can be generated in a so-called 'oddball task' (Fig. 2). In this task a train of stimuli is presented in which some stimuli are presented often (the 'standard' stimuli) and others rarely (the 'deviant' stimuli). When the deviants are attended to, a positive deflection is seen in the EEG about 300 ms after stimulus onset. This difference in brain activity between deviant and standard stimuli can be used in a BCI to identify which stimulus is attended to. Depending on which one, different computer commands can be executed.

Figure 2 Deviant target stimuli (T) within a sequence of standard stimuli (S)
can generate a P300 event-related potential (adapted from Polich, 2007)

In the P300 speller, the rows and columns of a letter matrix flash randomly and any letter in the matrix can be the deviant stimulus, simply by focusing on it: you could count or think 'yes!' when it flashes.

The video below shows the P300 speller BCI in our lab in action. The subject is spelling the words “HANDS FREE” and is already halfway:

What can you expect from this tutorial?

In this tutorial, we will cover the basic steps to successfully run a spelling session. You are not expected to have prior knowledge of how to mount and record EEG.

To proceed to the first part of the tutorial, go here: Mounting the EEG-net.

If you are interested in more detailed information, go to www.BCI2000.org:

References

Birbaumer, N. and Cohen, L.G. 2007. “Brain–computer interfaces: communication and restoration of movement in paralysis.” The Journal of Physiology 579(3):621 -636.

Farwell, L. A. and E. Donchin (1988). Talking off the top of your head: toward a mental prosthesis utilizing event-related brain potentials. Electroencephalography and clinical Neurophysiology 70(6): 510-523.

Nijholt, A., D. P.O Bos, and B. Reuderink. 2009. Entertainment Computing, 1, 85-94.

Polich, J. (2007). Updating P300: an integrative theory of P3a and P3b. Clinical Neurophysiology 118(10): 2128-2148.

Van Gerven, M., Farquhar, J., Schaefer, R., Vlek, R., Geuze, J., Nijholt, A., Ramsey, N., Haselager, P., Vuurpijl, L., Gielen, S. And Desain, P. 2009. The brain-computer interface cycle. Journal of Neural Engineering. 6, 1-10

Wolpaw, J.R., Birbaumer, N., McFarland, D.J., Pfurtscheller, G. and Vaughan, T.M. (2002). Brain-computer interfaces for communication and control. Clinical Neurophysiology, 113, 767-791

courses/brain-computer_interfaces.txt · Last modified: 2014/04/07 23:59 by Simon-Shlomo Poil
The NBTwiki platform - version 2.8 - 9 May 2013
Copyright (C) 2008-2015