Document Actions

You are here: Home / News / Network News / First Bernstein Symposium on Computational Neuroscience held in Freiburg

First Bernstein Symposium on Computational Neuroscience held in Freiburg

From 10 to 12 October, scientists from the four German Bernstein Centers for Computational Neuroscience (BCCNs) got together in Freiburg for their first symposium to present and discuss their latest research results. Last year, the BMBF (German Federal Ministry of Education and Research) had established the four Bernstein Centers to concentrate scientific expertise in neuroscience and thus get a step closer to the common goal of understanding thought (October 2005).

At this first ever Bernstein symposium, the scientists presented their contributions in various aspects of neuroscience. Common to all contributions, though, was the aim of increasing the use of mathematical or algorithmic models to understand the functioning of the brain. Over the last 10 years, these theoretical models have undergone dramatic progress. Whereas in the 1990s it had only been possible to computer-simulate individual nerve cells or small networks, it is possible these days to simulate networks of about 100,000 nerve cells with the help of specially developed software and state-of-the-art computer clusters. This is about the number of cells in one cubic millimetre of brain. The enormous technical challenges associated with such simulations can only be met in close cooperation with industrial partners (such as Honda Research Institute Europe GmbH, Infineon). However, despite the considerable progress it should not be forgotten that an actual human brain contains about 100,000 times the number of cells of such artificial networks. Even the most advanced of artificial networks are still a very long way behind the complexity of a biological brain.

Indeed, experimental results gained from actual brains keep baffling scientists with ever new unexpected characteristics. The Bernstein Symposium presented numerous such as yet unexplained phenomena; however, some of them may soon be understood using the newly developed models. The investigations comprised a wide range of organisms and behavioural patterns: from the common housefly to humans, from movement control to language recognition. It is striking to note that specific perceptive performances are realized differently in different organisms. Why is that the case? Andreas Herz (BCCN Berlin) put forward the hypothesis that the brain of each organism is structured in such a way as to ensure the best possible extraction of sensory information from the environment relevant to that organism .
Same as on this macroscopic level, there are many open questions on a microscopic level, too, for which scientists thought that they had come close to understanding. For example, there are still characteristics in the formation of individual nerve impulses (action potentials) that cannot be explained using the traditional textbook models (Fred Wolf and colleagues, BCCN Göttingen).

The practical importance of such findings in computational neuroscience lies particularly in biomedical applications. In Freiburg, for example, scientists in the group of Andreas Schulze-Bonhage work towards not only being able to predict epileptic seizures, but possibly being able to prevent them one day through electric stimulation (a kind of pacemaker for the brain). In the Bernstein Centers in Berlin, Freiburg and Göttingen, scientists work on the development of new procedures allowing paralyzed patients to control neuronal prostheses by their brain activity alone. Almost ready for practical implication is a language recognition system, developed by Munich scientists in cooperation with Infineon, which takes into account the latest physiological findings regarding the individual stages in the auditory pathway.

In addition to the many informative highlights, the symposium also debated aspects of strategy, such as ways to improve research and teaching in the field of Computational Neuroscience. Great importance was attributed to training programmes, particularly their national and international integration. To that end, representatives of comparable centers and training programmes in the UK, US, Israel and Japan were invited, together with those of US funding organizations. Undoubtedly, such intensive international coorporations are clearly the way for much future progress. The ultimate aim is for young neuroscientists not only to enjoy a specifically tailored interdisciplinary training programme, but to be able to complete such programme at various locations across the globe. In the Bernstein Centers, this vision of the future is almost within reach. International exchange agreements are already in place for some of the BCCN training programmes. In future, young Bernstein scientists may not have to wait for post-doc positions to travel the world, but do so during their doctoral studies. The expected scientific gain for doctoral projects and the widening of the personal horizon of the doctoral candidates involved are probably priceless.