Search this site

Organization

Bernstein Focus: Neurotechnology Berlin
 
Bernstein Center for Computational Neuroscience Berlin



© 2009 Berlin BCI

Programme

Wednesday, July 8, 2009
Morning Registration
09:45 - 10:00
 Welcome
 		Klaus-Robert Müller
    Session Chair: Steven Lemm
10:00 - 11:30 Tutorial
Electrophysiology		 Gabriel Curio
Warm-up for BCI physiology: Basic concepts in the transition from cellular microrecordings to noninvasive large-scale EEG/MEG signals
Macroscopic brain sigals, e.g., the noninvasively measured scalp EEG/MEG, correlate with mental states, such as movement intentions. This opens a new communication channel for paralyzed patients as their intentions can be 'read' by computers and utilised for triggering technical devices. This entry-level tutorial will provide an intuitive introduction into the relevant brain anatomy and physiology and describe typical procedures as well as potential pitfalls when obtaining and analysing multichannel EEG/MEG data.
11:30 - 11:45 Coffee Break
11:45 - 13:15 Tutorial
Non-Invasive		 Klaus-Robert Müller, Benjamin Blankertz
Machine Learning and Signal Processing Tools for BCI
We will first provide a brief overview of Brain-Computer Interface from a machine learning and signal processing perspective. In particular showing the wealth, the complexity and the difficulties of the data available, a truly enormous challenge: In real-time a multi-variate very strongly noise contaminated data stream is to be processed and neuroelectric activities are to be accurately differentiated. We will then in detail discuss the components of the data analysis chain employed in modern BCI systems, spanning all aspects from preprocessing and feature extraction, adaptive vs. fixed classification and feedback design.
13:15 - 14:45 Lunch Break
    Session Chair: Alexander Gail
14:45 - 16:15 Tutorial
Invasive		 Eilon Vaadia
About learning, predictions and adaptivity of brains and machines - Using useful signals from the brain, and useful computer algorithms to improve brain machine interfaces.
I will talk about the physiology of motor cortex and the nature of activity of population of neurons in motor cortex during Sensorimotor learning, movement preparation and execution. I will present the approach of internal models of the brain as the basis for learning and perception and use all of the above to show how our current knowledge can facilitate approaches to adaptive brain machine interfaces.
16:15 - 16:30 Coffee Break
16:30 - 18:00 Tutorial
Invasive		 Gerwin Schalk
Theory and Application of Electrocorticographic (ECoG) Signals in Humans
Brain-computer interfaces (BCIs) convert brain signals into outputs that communicate a user's intent. BCIs can be used by people to communicate and interact with their environment.  However, the prevailing non-invasive and invasive sensor methods have important limitations.  Electrocorticographic (ECoG) recordings from the surface of the brain could be a robust and high-fidelity alternative to existing sensor methods.  This tutorial will provide an overview of the history of ECoG recordings; describe the types of signals present in ECoG and their relationship to signals detected using EEG and intracortical microelectrodes; and finally give examples of successful use of these signals in real time for BCI purposes and also for diagnosis. [more]
18:00 - OPEN Poster Session

Evening BBCI Barbecue

Thursday, July 9, 2009
    Session Chair: Eilon Vaadia
09:00 - 09:40 Workshop
Invasive		 Alexander Gail
What's the plan? - Movement-Goal representations in the frontoparietal reach network
When planning goal-directed arm movements the sensorimotor system needs to integrate the current behavioral context with given spatial constraints to define and maintain motor-goals. We are interested in how this integration is achieved within the sensorimotor network. Specifically we tested the respective roles of the dorsal premotor cortex (PMd) and the parietal reach region (PRR) in defining and planning context-specific reach goals. [more]
09:40 - 10:20 Workshop
Invasive		 Mikhail Lebedev
Brain-machine interfaces based on neuronal ensemble recordings 
Brain-machine interfaces (BMIs) have experienced an explosive development during the last decade. Current state of the art BMIs convert neuronal ensemble activity recorded in nonhuman primates or human subjects into reaching and grasping movements performed by artificial actuators. BMIs that enact movements of lower extremeties are less explored. Additionally, most BMI implementations do not have somatosensory feedback from the actuator. I will review our recent experiments in which we (i) extracted bipedal locomotion patterns from monkey cortical activity and (ii) used spatiotemporal patterns of intracortical microstimulation to deliver information back to the brain. These results bring us closer to building clinical neuroprosthetic devices for restoration of both sensory and motor functions in paralyzed people.
10:20 - 11:00 Workshop
Invasive		 Andrew Jackson
Plasticity at the Brain-Computer Interface
Next generation recurrent Brain-Computer Interfaces will not only extract signals from cortical activity but also deliver feedback to the nervous system via electrical stimulation. For example, stimulation of cervical spinal segments can produce functional arm and hand movements such as reaching and grasping. We are developing new technologies including chronic electrodes and implantable electronic circuitry to control stimulation from cortical recordings, constituting an artificial corticospinal connection which could replace injured motor pathways. I will present evidence that the motor system can readily acquire the novel neuromotor transformations required to incorporate these connections into motor system function. In separate experiments we have shown that operation of artificial connections can potentiate new motor pathways via activity-dependent plasticity mechanisms. Together, these results suggest that recurrent BCIs have application not only as prostheses to replace function, but also as tools for manipulating plastic reorganisation to restore nervous system function following injury. [more]
11:00 - 11:30 Coffee Break
    Session Chair: Gabriel Curio
11:30 - 12:15 Workshop
Invasive		 Niels Birbaumer
BCI in paralysis: an unfulfilled promise
EEG and ECoG (Electrocorticogram) can be used successfully to initiate direct brain communication with locked-in patients but fail in completely locked-in patients. Possible reasons are explained and some new solutions with first data presented. In chronic stroke the author's team together with L. Cohen's group at NIH have shown motor restoration of paralysed hand in chronic stroke without residual movement using non-invasive MEG/BCI. However generalization from the BCI-clinic to the social reality was poor. A new strategy for invasive and non-invasive BCI in chronic stroke is demonstrated and first data presented. [more]
12:15 - 13:00 Workshop
EEG		 Gert Pfurtscheller
The Hybrid BCI
There are several different BCI approaches, which may or may not depend on external stimulation. Slow cortical potential (SCP)-, event-related desynchronisation (ERD)- and sensorimotor rhythm (SMR)-BCIs do not require external stimulation, while P300-BCIs and steady state visual evoked potential (SSVEP)-BCIs do. Dependent means in this respect that the user has to focus attention and/or gaze to flickering/flashing lights or and can therefore not completely freely decide to perform an action. Each type of BCI system has advantages and disadvantages. SSVEP-BCIs need minimal training time and can achieve a high information transfer rate (ITR), but have a relatively high false positive rate during rest. In contrast, an asynchronous brain switch based on the post-imagery beta ERS has a low ITR, but can be set up quickly and easily with a low false position rate (Pfurtscheller and Solis-Escalante, Clin. Neurophysio. 2009). It is therefore a challenge to use the advantages of different BCI systems and create a “hybrid” BCI system by switching e.g. a battery of flickering lights (SSVEP-BCI) on or off by using a brain switch (ERD-BCI). Another type of “hybrid” BCI can analyse motor imagery related EEG changes and SSVEP amplitudes simultaneously. It was shown recently that such a “hybrid” strategy results in a better classification accuracy relative to either an ERD or SSVEP classification alone (Allison et al submitted 2009). [more]
13:00 - 14:15 Lunch Break
    Session Chair: Andrea Kübler
14:15 - 15:00 Workshop
EEG		 Theresa Vaughan
EEG‐based Brain‐Computer Interface for Communication and Control: Independent Home Use
People affected by severe motor disorders such as amyotrophic lateral sclerosis (ALS), brainstem stroke, cerebral palsy, and spinal cord injury need alternative methods for communication and control. They may not be able to use even the most basic conventional assistive technologies, which all rely in one way or another on muscles. Studies from this and other laboratories have shown that humans, including those with severe motor disabilities, can learn to control sensorimotor rhythms and other features of scalp‐recorded electroencephalographic (EEG) activity and that they can use this control to select letters or icons, or move a cursor in up to three dimensions. Such multidimensional control could be used to control a prosthesis or a robotic arm. Currently, we are showing that people with ALS can use EEG‐based brain‐computer interfaces (BCIs) for communication and control independently in their homes. [more]
15:00 - 15:45 Workshop
EEG		 Christa Neuper
Feedback-regulated mental imagery in BCI applications: using non-invasive EEG and NIRS signals
An important issue of brain-computer interface (BCI) development is to detect changes in brain signals that are related to specific intentions or thought processes. For example, mental motor imagery modulates the sensorimotor brain activity, and the detected changes can be used to operate a computer-controlled device. Clinical applications of this technology include the restoration of movement, such as control of grasping with the help of a neuroprosthesis, in severely paralyzed individuals. The motor imagery based BCI training may further be useful as a complementary therapeutic tool to facilitate functional recovery after stroke. To date, the majority of BCI systems rely on EEG recordings. However, near-infrared spectroscopy (NIRS) has recently attracted attention of BCI researchers due to its noninvasiveness, portability, short preparation time, and relatively low cost. In this talk I will shortly introduce the NIRS technique for BCI development and present data on how characteristic hemodynamic responses during motor imagery can be modulated by real-time NIRS feedback. Based on recent results I will finally discuss how simultaneous NIRS and EEG recordings might combine advantages of both approaches. [more]
15:45 - 16:00 Coffee Break
    Session Chair: Benjamin Blankertz
16:00 - 16:45 Workshop
EEG		 Michael Tangermann
Tackling increasing opthalmologic problems in patients with a new auditory multi-class BCI paradigm
Most P300 BCI approaches use the visual modality for stimulation and feedback. Due to increasing sight deterioration, this might not be the preferable choice for ALS patients in late stages. To tackle this general problem, a multi-class brain-computer interface paradigm is proposed that uses fast spatially distributed auditory cues for ERP paradigms.
16:45 - 17:30 Workshop
EEG		 Andrea Kübler
BCI communication and beyond: Psychological influences and new perspectives
It has often been proposed that psychological factors influence BCI performance. I will present data about how emotional state and motivation may affect BCI performance obtained from healthy volunteers and those with neurological disease. Additionally I will present more application oriented data on how BCI can be used for creative expression and the influence on quality of life. From this complex form of BCI use I will go back to the roots of basic communication with BCI and the possible application of BCIs for diagnosis in non-responsive patients.
17:30 - OPEN Poster Session

17:30 - 18:30 Event / Press

 Brain Painting Event
Press Background Talks

Friday, July 10, 2009
    Session Chair: John-Dylan Haynes
09:00 - 09:45 Workshop
Signals		 Nikos K. Logothetis
Neurovascular Coupling: Insights from Physiology, Neuropharmacology & Electrical Microstimulation
I shall describe our current understanding of the neurophysiological and hemodynamic signals, and of the functional neurovascular coupling in the anesthetized and alert behaving monkey. The neurovascular coupling was studied by means of physiological and fMRI experiments, during neuropharmacological blocking of pyramidal cell activity, and with combined electrical microstimulation and fMRI (esfMRI).
09:45 - 10:30 Workshop
fMRI		 Rainer Goebel
A robust spelling device for locked-in patients based on real-time fMRI
Several medical conditions (e.g., brain injury, stroke, progressive neurological diseases) can lead to complete paralysis while largely preserving sensory and cognitive functions and associated brain activation. We investigated whether healthy subjects are able to "write" solely on the basis of voluntary control of the fMRI (BOLD) signal. Using a guided display technique, we show that subjects can learn in less than half an hour to produce reliably any letter of the alphabet in a single trial. To achieve this performance, subjects use three mental strategies to modulate spatio-temporal properties of the fMRI signal in three different brain areas. While the transmitted information (BOLD time courses from regions-of-interest) has been initially decoded offline by human raters, we have recently implemented a fully automatized real-time "brain reading" technique.The developed paradigm and decoding technique might be applied in locked-in patients to let them communicate their wishes and thoughts reliably without extensive pre-training.
10:30 - 10:45 Coffee Break
    Session Chair: Klaus-Robert Müller
10:45 - 11:25 Workshop
fMRI		 John-Dylan Haynes
FMRI-decoding of mental states from brain activity:From basic science to applied neurotechnology
Recent advances in neuroimaging have shown that it is possible to decode a person's mental states with high precision. Such "brain reading" is achieved by training pattern classifiers to recognize the specific activity patterns that encode individual thoughts in the human brain. Many types of mental states have been decoded, including percepts, memories, intentions, and even unconscious decisions. This research has already provided new insights into a number of fundamental questions regarding the human mind, such as the link between brain and conscious experience, the processes underlying decisions and free will. Now this field is also beginning to explore the first promising applications. The list of potential real-world applications is long and includes brain-computer-interfaces, detection of residual mental activity in coma patients, lie detection, reading out concealed mental states and neuromarketing. This talk will give an overview of this emerging field with a particular emphasis on potential applications.
11:25 - 12:05 Workshop
fMRI		 Jack Gallant
Voxel-based encoding models in fMRI and their use for decoding and BCI
Most current work on brain-computer interfaces focuses on decoding algorithms that aim to classify, identify or reconstruct the stimulus directly from measured brain activity. These approaches are often based on non-parametric algorithms such as support vector classification that do not include any explicit information about how the brain encodes information. My laboratory has pioneered an alternative approach that is based on estimating explicit non-linear encoding models that describe how stimuli are transformed into measured brain activity. These models can be evaluated in the forward direction to investigate how the brain represents visual information. Furthermore, they can be inverted and combined with an appropriate prior to obtain Bayesian decoding models that can be used to reconstruct perceptual experiences. This Bayesian decoding framework is quite general: in theory it can be applied to any brain activity measurements, and can be used to reconstruct many perceptual and cognitive states. Thus, Bayesian decoding algorithms might form the basis of powerful new brain-reading technologies and brain-computer interfaces. [more]
12:05 - 12:45 Workshop
fMRI		 Lars Kai Hansen
Modeling fMRI dynamics
Functional MRI modeling is challenged by long-range coupling, non-linearity, and lack of detailed physiological information. I will review our progress in modeling the spatio-temporal dynamics of fMRI including hemodynamic deconvolution, blind deconvolution, and brain state decoding based on spatio-temporal kernel methods.
12:45 - 14:15 Lunch Break
    Session Chair: Ulrich Egert
14:15 - 15:00 Workshop Wilhelm Kincses, Klaus-Robert Müller
Brain@Work - Neurotechnology-based Man-Machine Interaction for Industrial Applications
15:00 - 15:45 Workshop

 Andrea Kübler, Donatella Mattia, Rüdiger Rupp, Gernot Müller-Putz 
TOBI: Tools for Brain-Computer Interaction
The TOBI project aims to develop practical technology for brain-computer interaction; i.e., non-invasive BCI prototypes combined with other assistive technologies (AT) that will have a real impact in improving the quality of life of disabled people. These non-invasive BCI are based on electroencephalogram (EEG) signals. Researchers will be looking at a hybrid approach whereby users can fuse brain interaction and muscle-based interaction or can switch between different channels naturally (based on monitoring of physiological parameters or mental states).
15:45 - 16:00 Coffee Break
    Session Chair: Michael Tangermann
16:00 - 16:45 Workshop
Neurotechnology		 Klaus-Robert Müller, John-Dylan Haynes, Sebastian Möller
Bernstein Focus: Neurotechnology Berlin
The Bernstein Focus: Neurotechnology Berlin (BFNT-B) posits that neuroscientific results can be exploited for developing robust ‘real-world’ applications that have a major potential for (also non-medical) industry. Similar to the new paradigm of medical research ‘from bench to bedside and back’, the center brings together a multidisciplinary faculty with the aim of directly applying insights from basic neuroscience to relevant applications (‘from bench to desktop and back’). The major aim of the BFNT-B is to foster novel noninvasive ‘brain reading’ techniques to enhance man-machine interactions. Their contributions will be evaluated, e.g. in the future-oriented field of usability studies for telecommunications systems and services, or driver-assisted measures for vehicle safety.
16:45 - 17:30 Workshop
Neurotechnology		 Carsten Mehring, Ulrich Egert
Bernstein Focus: Neurotechnology Freiburg
In spite of considerable progress towards prosthetic devices controlled by neuronal signals, brain-machine interfaces and other neurotechnological devices, however, user-friendly, neurotechnical devices for everyday use remain a vision of the future, with numerous fundamental biological, technical, computational, clinical, and ethical problems still to be solved. The aim of the BCNT-FT consortium is the development of bidirectional hybrid neurotechnical devices for human usage. This will be implemented in three research clusters, each consisting of projects organized around a common goal: to understand the principles, to advance technology, and to explore and extend clinical applications. Each project addresses issues central to neurotechnology, from basic questions on decoding neuronal signals, interfacing biological neuronal networks to technical devices, actuators and real-time feedback systems, via the stable recording and interpretation of neuronal signals, to the clinical testing and application of new technologies. Research in the BCNT-FT will be supported by a matching, interdisciplinary training program for neurotechnology. In collaboration with industrial, applied and clinical partners neuroprosthetical devices for biomedical application will be developed.
17:30 - 17:45 Goodbye
Closing Remarks		 Presentation of the "Best Poster Award".
19:00 Speakers Dinner

Attachments (11)

  • Abstract BBCI Berlin Schalk.pdf - on Apr 27, 2009 12:03 PM by Matthias Jugel (version 1)
    14k View Download
  • BBCI 09 Abstract Eberhard E. Fetz.pdf - on Apr 27, 2009 11:43 AM by Matthias Jugel (version 1)
    50k View Download
  • BBCI 2009 Birbaumer.pdf - on Apr 29, 2009 1:02 AM by Matthias Jugel (version 1)
    22k View Download
  • Berlin July o9 - Pfurtscheller.pdf - on May 18, 2009 3:08 AM by Matthias Jugel (version 1)
    57k View Download
  • Berlin_BCI_abstract_Jackson.pdf - on Jun 25, 2009 3:06 AM by Matthias Jugel (version 1)
    38k View Download
  • Computing with Neural Ensembles abstract_2009_Berlin.pdf - on Apr 27, 2009 12:00 PM by Matthias Jugel (version 1)
    8k View Download
  • Gail 2009 BBCI Berlin.pdf - on Apr 28, 2009 11:05 PM by Matthias Jugel (version 1)
    31k View Download
  • Gallant.BBCI09.Abstract.pdf - on May 4, 2009 6:00 AM by Matthias Jugel (version 1)
    36k View Download
  • Neuper_Berlin09_BCI.pdf - on May 19, 2009 11:23 AM by Matthias Jugel (version 1)
    6k View Download
  • Vaadia BBCI 2009.pdf - on May 22, 2009 3:21 AM by Matthias Jugel (version 2 / earlier versions)
    31k View Download
  • Vaughanetal_Berlin09.pdf - on Apr 27, 2009 11:39 AM by Matthias Jugel (version 1)
    87k View Download