Metabolic brain-computer interfaces [Elektronische Ressource] / vorgelegt von Ranganatha Sitaram

Ranganatha Sitaram

Sujets

• Gesundheit

Informations

• Publié par: eberhard_karls_universitat_tubingen
• Publié le: 01 janvier 2010
• Nombre de lectures: 14
• Langue: English
• Poids de l'ouvrage: 8 Mo

Extrait

Aus dem Institut für Medizinische Psychologie
der Universität Tübingen
Direktor: Professor Dr. N. Birbaumer


Metabolic Brain-Computer Interfaces


Inaugural-Dissertation
zur Erlangung des Doktorgrades
der Humanwissenschaften

der Medizinischen Fakultät
der Eberhard-Karls-Universität
zu Tübingen

vorgelegt von
Ranganatha Sitaram
aus
Harihar/Indien

2010























Dekan: Professor Dr. I. B. Auterieth

1. Berichterstatter: Professor Dr. N. Birbaumer
2. Berichterstatter: Professor Dr. B. Pichler
2












To my mother and late father.
3
Acknowledgements
First and foremost, I would like to express my deepest gratitude to Prof. Dr. Niels
Birbaumer who was instrumental in my moving to Tuebingen 4 years ago and thus
enabled me to do this research. From the day in early 2004 in Singapore when I first
expressed my interest to work with him to this day his kindness, ready wit and a
certain zeal for scientific investigation has inspired and motivated me. Without his
characteristic generosity, openness to ideas and freedom afforded by him at work my
move to research in cognitive neuroscience would never have been possible. I’m also
ever so grateful to the secretary Angela Straub whose help and assistance made my
family’s stay in Tuebingen possible.

My heartiest thanks to my second supervisor and former director Dr. Guan Cuntai
who with his kind countenance not only encouraged me to pursue my interests but
also actively collaborated with me in this research.

I’m at a loss for words to acknowledge the support, fortitude, reassurance, stimulus
and solace that I found in my wife, Vani, and children, Nayantara and Anirudh during
this period.

I`m indebted to my colleagues Andrea Caria, Ralf Veit and Giuseppina Rota who
were my constant companions in many of our joint experiments, and who enabled me
to appreciate the good, bad and ugly of experimental neuroimaging. But for their
composure, good humour and loving kindness, I would not have the feeling of a
delectable taste that the completion of this work has metaphorically left behind in my
mouth.

I would also like to convey my gratefulness to the immense technical help and a
number of fruitful discussions during my fMRI measurements from Dr. Michael Erb
and Prof. Dr. Wolfgang Grodd from the Neuroradiology Department, University
Kilinkum, Tuebingen.

My special thanks to close friends and research partners, Sangkyun Lee, Sergio Ruiz,
Balint Varkuti, Ahmed Karim, Boris Kleber and Massimiliano Rea for the many
stimulating discussions, exciting projects and grant proposals that I had the privilege
to be involved in. The convivial atmosphere they created has made this work the most
enjoyable in my career.

Grants from the SFB 437 “Kriegserfahrungen” and SFB 550 of the Deutsche
Forschungsgemeinschaft (DFG) supported this work.






4
Contents

Introduction....................................................................................................................7
Functional Near Infrared Spectroscopy.......11
Introduction..............11
Basic Theory of NIRS..............................................................................................11
Interaction of light with brain tissue....................................11
Physiological changes during brain activity........................12
NIRS signal acquisition techniques .........................................13
NIRS signal analysis ................................14
Benefits and limitations of NIRS.............................................14
Functional imaging with NIRS................14
BCI development with fNIRS..................................................15
Chapter 2......................................................................................18
Implementation of an fNIRS Brain-Computer Interface.............18
Introduction..............18
Materials and Methods.............................................................20
Subjects ................................................................................20
Experimental Procedure.......................21
Signal Acquisition....................................22
Preliminary Signal Analysis ....................................................22
Pattern Classification...............................24
Support Vector Machine (SVM)..........................................24
Hidden Markov Model.........................27
Graphical user interface of the NIRS-BCI system...................28
Results......................................................................................................................30
Discussion................34
Chapter 337
Functional Magnetic Resonance Imaging Brain-Computer Interface .........................37
Introduction..............................................................................................................37
Architecture of fMRI-BCI.......................37
Signal Acquisition................................................................................................39
Signal Preprocessing............................40
Signal Analysis....42
Signal Feedback...45
Preliminary Investigations .......................................................................................48
Regulation of Anterior Insula in Healthy Individuals..........48
Regulation of Ventrolateral Premotor Cortex in Healthy Individuals & Stroke
Patients.................................................51
Self-regulation of Right Inferior Frontal Gyrus and Language Processing.........57
Chapter 4......................................................................................62
Decoding Emotion States of the Brain from fMRI Signals.........62
Introduction..............................................62
Method.....................64
Results......................................................................................67
Discussion................68
Chapter 5................................69
Volitional Regulation of Anterior Insula in Healthy Individuals and Psychopathic
Criminals......................................................................................69
Introduction..............69
5
Neurobiology of Insula ............................................................................................69
Psychopathy .............................................70
Study Objectives......................................72
Methods....................72
Participants...........................................72
fMRI Data Acquisition........................................................73
Experimental Protocol .........................................................73
Off-line Data Analysis75
Results and Discussion............................77
Experiment 1: Healthy individuals ......................................................................77
Experiment 1: Discussion....................85
Experiment 2: Results ..........................................................85
Experiment 2: Discussion................................91
Chapter 6......................................................93
Discussion....................93
Hemo-neural Model of Regulation ..........................................................................94
References97

6
Introduction

Brain-Computer Interfaces (BCI) utilise neurophysiological signals originating in the
brain to activate or deactivate external devices or computers (Donoghue 2002;
Wolpaw, Birbaumer et al. 2002; Nicolelis 2003; Birbaumer and Cohen 2007). The
neuronal signals can be recorded from inside the brain (invasive BCIs) or outside
(non-invasive BCIs) of the brain. Most BCIs developed so far have used operant
training of direct neuroelectric responses, Electroencephalography (EEG) waves,
event-related potentials and brain oscillations (Birbaumer, Weber et al. 2006;
Birbaumer and Cohen 2007). Compared to neuroelectric studies on regulation of brain
activity, there have been fewer studies with metabolic signals from the brain (Sitaram,
Caria et al. 2007; Weiskopf, Sitaram et al. 2007; Sitaram, Weiskopf et al. 2008). Near
Infrared Spectroscopy (NIRS) and Functional magnetic resonance imaging (fMRI)
present themselves as attractive methods of acquiring hemodynamic activity of the
brain for a developing a BCI. In this study, we exploit NIRS and fMRI for the
implementation of BCIs for the investigation of regulation of hemodynamic signals in
the brain and their behavioural consequences. We propose that these methods could
be used not only for communication and control in paralysis, but also as powerful
tools for experiments in neuroscience and rehabilitation and treatment of neurological
disorders.


Research on operant training to self-regulate brain responses began with early studies