The Eighth IASTED International Conference on
Signal Processing, Pattern Recognition, and Applications
SPPRA 2011

February 16 – 18, 2011
Innsbruck, Austria

TUTORIAL SESSION

Signal Processing and Pattern Recognition Methods in Brain-Computer Interfaces

Dr. Gary Garcia-Molina
Philips Research, North America, USA
gary.garcia@philips.com

Abstract

fiogf49gjkf0d
Research in Electroencephalogram (EEG) based Brain-Computer Interfaces (BCIs) has been considerably expanding during the last few years. Such an expansion owes to a large extent to the multidisciplinary and challenging nature of BCI research. Signal processing and pattern recognition undoubtedly constitute essential components of a BCI system. Signal processing algorithms are applied to the EEG signals to extract features that characterize the mental activities utilized in the BCI system. These features are grouped into feature vectors which are translated into commands by a classifier.
In this tutorial, the basic BCI concepts, brain activity monitoring, BCI operation, and the electrophysiological sources of BCI control, are introduced. The main BCI types, namely motor imagery (ERD/ERS), steady state visual evoked potentials (SSVEP), and P300 based BCIs are presented along with practical application examples.
The EEG processing for BCI applications is then described in depth. The multivariate nature of the EEG combined with the neuroscience knowledge on hemispheric brain specialization are advantageously taken into account to derive spatial filters (i.e. across the EEG electrodes) to obtain feature vectors resulting from motor imagery, visual evoked potentials, and the P300 paradigm.
The machine learning approaches to translate the feature vectors into commands to control an application are reviewed. A special emphasis is put on the comparison between linear and nonlinear classifiers. Most of current research results support the use of linear classifiers because an easy link can be found between the classifier structure and neurophysiology. This view is discussed in the tutorial and examples are provided where a nonlinear classifier is more suitable.
Throughout the tutorial, the applicability of BCI technology is emphasized. More than two decades of BCI research have been paving the way to the deployment of BCIs from the lab to home.

Objectives

fiogf49gjkf0d
The objectives of this tutorial lecture are:

Timeline

fiogf49gjkf0d
Part 1. Brain-Computer Interfaces (30 minutes)
This part of the tutorial covers general BCI notions and proposes answers to a number of commonly asked questions on BCIs.
Part 2. Main BCI types (30 minutes)
BCIs can use a variety of electrophysiological sources. However, more than 90% of current BCI implementations rely on three main electrophysiological sources. These are presented in this part of the tutorial.
Part 3. EEG processing and Machine learning for BCI applications (2 X 45 minutes)
In this part of the lecture, EEG processing for BCI is discussed in depth. A common framework to address the detection of patterns associated with the three main BCI types is presented.
The tutorial ends with a brief overview of the open research questions and a roadmap towards the mainstream adoption of BCI technology.

Target Audience

fiogf49gjkf0d
The subjects in this tutorial are comprehensibly presented and are thus suitable for a wide range of scientists. The approaches, methods, and algorithms in this tutorial do not limit to the BCI realm. Indeed, they can be readily applied to domains where other physiological signals are used.
BCI is a multidisciplinary field comprising neurophysiology, signal processing, machine learning, user interaction, and information theory. This tutorial is prepared in such a way as to clearly introduce each concept and its relevance for BCI.

Qualifications of the Instructor(s)

fiogf49gjkf0d
Dr. Gary Garcia Molina has been active in BCI research for a decade.
In 2004, he obtained his doctoral degree from the Swiss Federal
Institute of Technology Lausanne, Switzerland (EPFL). His thesis
entitled "Direct Brain-Computer Communication Through Scalp Recorded
EEG Signals" was nominated for the Swiss best thesis award.
Since January 2005, Dr. Garcia is a Senior Researcher at Philips
Research Europe Laboratories (Eindhoven, The Netherlands) where he
leads a research activity aiming at developing practical
Brain-Computer Interfaces for the consumer market. From 2007, Gary
Garcia leads the Philips team that participates in the European
consortium BRAIN (www.brain-project.org) which has as main objective
the development of BCIs able to automatically adapt to the user and
his/her environment, and do not require any expert assistance.
Gary Garcia serves as an adviser and project reviewer for the
European-research program ICT (Information and Communication
Technologies) on Inclusion and Independent Living.
Dr. Garcia published numerous research papers, and holds four
patents on BCI related technology. He has extensive experience with
several types of BCIs and has been cooperating with numerous
laboratories academic and industry to develop BCI technology into
home appliances for the disabled and the healthy. He chaired
conference sessions on signal processing for BCI technology, and led
Philips seminars on EEG processing. In November 2008, he taught a
tutorial on BCIs based on Motor Imagery at the IEEE/BiOCAS
(Intelligent Biomedical systems) conference
(http://www.biocas2008.org/tutorials.html). In September 2009 he
gave a tutorial on Affective BCIs~\cite{garciaTutorialACII} at the
"Affective Computing and Intelligent Interaction" (ACII) 2009
conference. In 2010 he gave two tutorials on signal processing for
EEG signals and BCI applications at the "European Signal Processing
Conference" EUSIPCO 2010, and the "International Conference on
Information Science, Signal Processing, and their Applications
(ISSPA 2010)".

References

[1] 
fiogf49gjkf0d
G. Garcia Molina, T. Ebrahimi, and J.-M. Vesin. Joint Time-Frequency-Space Classification of EEG in a Brain-Computer Interface Application. EURASIP Journal on Applied Signal Processing, 2003(7):713–729, 2003.
[2] G. Garcia Molina. Electroencephalogram processing in Motor Imagery Based BCI: A tutorial.
[3] http://www.biocas2008.org/tutorials.html, 2008.
[4] G. Garcia Molina. Direct Brain-Computer Communication Trough Scalp recorded EEG Signals. PhD thesis, Swiss Federal Institute of Technology, Lausanne, 2004.
[5] G. Garcia Molina. BCI adaptation using incremental-SVM learning. In IEEE-EMBS International Conference on Neural Enginnering (NER), 2007.
[6] G. Garcia Molina. Spatial filters for detecting Steady State Visual Evoked Potentials: BCI application. In NeuroMath Workshop Leuven, 2009.
[7] G. Garcia Molina. Detection of High-Frequency Steady State Visual Evoked Potentials Using Phase Rectified Reconstruction. In 16th European Signal Processing Conference EUSIPCO 2008, 2008.
[8] G. Garcia Molina. High frequency SSVEPs for BCI applications. In Workshop at CHI2008: Brain-Computer Interfaces for HCI and Games, 2007.
[9] U. Hoffmann, G. Garcia Molina, J.-M. Vesin, and T. Ebrahimi. Application of the evidence framework to brain-computer interfaces. In IEEE Engineering in Medicine and Biology Society Conference, 2004.
[10] U. Hoffmann, G. Garcia Molina, J.-M. Vesin, K. Diserens, and T. Ebrahimi. A boosting approach to p300 detection with application to brain-computer interfaces. In IEEE EMBS Conference on Neural Engineering, 2005, 2005.
[11] G. Garcia Molina, T. Tsoneva, and A. Nijholt. Tutorial on Emotional Brain-Computer Interfaces.
[12] http://www.acii2009.nl/program/show slot/9, 2009.
[13] T. Ebrahimi, J.-M. Vesin, and G. Garcia Molina. Brain-computer interface in multimedia communication. IEEE Signal Processing Magazine, 20(1):14–24, 2003.
[14] G. Garcia Molina, F. Bruekers, M. Damstra, and H. Weda. Logging brain signals. European Patent Filing 071178586, 2007.
[15] D. Zhu, J. Bieger, G. Garcia Molina, and R.M. Aarts. A survey of stimulation methods used in SSVEP-based BCIs. Journal of Computational Intelligence and Neuroscience, In Press, 2010.
[16] G. Garcia-Molina and V. Mihajlovic. Spatial filters to detect Steady State Visual Evoked Potentials elicited by high frequency stimulation: BCI application. Journal of Biomedizinische Technik / Biomedical Engineering, 55(3):173–182, 2010.