Event Calendar:

August 2014

Mo Tu We Th Fr Sa Su
31 28 29 30 31 01 02 03
32 04 05 06 07 08 09 10
33 11 12 13 14 15 16 17
34 18 19 20 21 22 23 24
35 25 26 27 28 29 30 31

» All Events

Project Leaders

Dr.-Ing. Moritz Grosse-Wentrup
Phone: +49 7071 601-542
Fax: +49 7071 601-552
moritz.grosse-wentrup[at]tuebingen.mpg.de
Prof. Dr. Bernhard Schölkopf
Phone: +49 7071 601-551
Fax: +49 7071 601-552
bernhard.schoelkopf[at]tuebingen.mpg.de

 

Machine Learning in Neuroscience

The neurosciences present some of the steepest challenges to machine learning. Among diverse problem settings and approaches, certain commonalities can be identified. Nearly always there is a very high-dimensional input structure - particularly relative to the number of exemplars, since each data point is usually gathered at a high cost in time and money. To avoid overfitted solutions, inference must therefore make considerable use of domain knowledge from physics, neurophysiology and anatomy. Solutions typically occupy a relatively small subspace of the input representation, the rest being made up of noise that may be of much larger magnitude (often composed largely of the manifestations of other neurophysiological processes, besides the ones of interest). In finding generalizable solutions, one usually has to contend with a high degree of variability, both between individuals and across time, leading to problems of covariate shift and non-stationarity. In all cases, even the high-dimensional raw input is a vastly simplified reflection of the underlying processes and structures. New ways of measuring relevant information, and new ways of transforming the data, are therefore still waiting to be found, leading to feature representations that are more relevant, less noisy, or more transferrable between experimental sessions and subjects.

Brain-Computer Interfacing for Communication

One specific neuroscientific application area in which we have an active interest is that of brain-computer interfacing (BCI). This research primarily aims to construct systems that could allow a paralyzed person to communicate, by decoding the user's intentions from measured brain signals. Currently, BCI systems based on electroencephalogram (EEG) signals do allow communication, and can be used even by people with very little remaining motor control, but they are still slow and difficult enough to use that they are not an attractive alternative to other communication methods. Among people who have absolutely no voluntary motor control, and who therefore stand to benefit most from BCI since they have no alternative, a convincing demonstration of successful communication has yet to be published. We have been working hands-on in collaboration with two Tübingen University departments, the Institute of Medical Psychology and Behavioral Neurobiology, and the Department of Neurosurgery at the University Clinics, in order to develop and test BCI systems with paralyzed patients. In addition, we pursue several laboratory-based lines of research. The contribution of machine learning to BCI is in developing, refining and applying algorithms to improve the accuracy with which neural signals are decoded, to interpret users' communication intentions more reliably and to reduce training times for the user. Methods with good generalization properties may especiallys improve performance in the hardest cases, where data sets are small or particularly noisy, as is typically the case with patient data. We regard algorithmic development and improvements in experimental methodology as significant contributions towards making clinical BCI systems a reality.

Brain-Computer Interfacing for Rehabilitation

While BCI is primarily regarded as an alternative means of communication for people with no voluntary motor control, subject groups with diverse cognitive deficits may also benefit from this line of research. BCI-technology can provide subjects with feedback on their neural signals, which may be utilized to learn how to volitionally induce beneficial brain states. One of our contributions to this field is the development of BCI-technology for the control of robotic devices, which we pursue in collaboration with our robotics laboratory. The combination of robotic devices with BCI-technology may be of particular value in the domain of stroke rehabilitation, as BCI-control of robotic-based physical therapy may result in enhanced rehabilitation outcomes. We evaluate the viability of such approaches in collaboration with the Department of Neurosurgery at the University Clinics.

Brain Connectivity

In contrast to the engineering approach of BCI, another role that machine learning can play in neuroscientific research is that of analysis-by-synthesis. We pursue a particular interest in causal inference problems that arise in the context neuroscientific research. Functional brain networks are highly recurrent due to their local organization in microcircuits and the bidirectional connections linking remote cortical areas. To understand how information processing is distributed in these structures, we are developing new tools in collaboration with the research group on causal inference at the MPI. In particular, we develop information-theoretical techniques for geometrically representing causal interactions, and kernel methods for detecting statistical dependencies in experimental time series. We apply these methodologies to neurophysiological recordings in order to provide new insight into the mechanisms of perception, and allow computational theories of these mechanisms to be advanced.
Last updated: Friday, 14.01.2011