Uni-Logo
You are here: Home Partners Visiting Scientists George Gerstein at the BCCN Freiburg
Document Actions

George Gerstein at the BCCN Freiburg

George Gerstein at the BCCN Freiburg

gerstein2.jpg From 10 September to 20 December 2006 I had the pleasure of making a research visit to the BCCN Freiburg.

This visit was an opportunity to connect again with old friends (Ad Aertsen, Markus Diesmann, Stefan Rotter and Uli Egert) and to get to know and work with a number of new ones. The day after my arrival we all went off to the ‘hut’ seminar, so I was able to hear about the various ongoing BCCN projects, and of course to meet a number of the people. I had arrived directly from a two week workshop on spike train analysis held in Newcastle, UK and organized by Stuart Baker.
Initially then, after the ‘hut’ seminar, I continued working on two topics that I had started at that workshop, one associated with finding neuronal groupings (like synfire chains), and the other with detecting significantly repeating spike patterns (‘melodies’) in simultaneous recordings of several neurons from cortex of behaving monkeys. This was largely a programming and algorithm effort, and resulted in both cases in large improvement in speed during production. The work with repeating patterns involved very large data arrays, and I was glad to have the opportunity to use the BCCN’s LOKI computer cluster with its enormous memory.

A year earlier Markus Diesmann had shown me the peculiar vertical stripe texture of spike time rasters from large scale network simulations. Data shown at the ‘hut’ seminar from synfire net simulations suggested to me that such striping originated in the time varying simultaneous activity of the several synfire chains. Together with the PhD student Sven Schrader and (later) Markus Diesmann we began to investigate this idea, and quickly generalized it to a new measure that compares which neurons are active at any two times. This new measure turns out to show a specific signature that detects the activity of synfire chains. It is remarkably sensitive, showing clear detection even if we sample only 200 neurons out the 50,000 in the simulated network with embedded synfire chains. Even though I have now returned to Philadelphia, we continue to study the new activity measure as applied to simulations of various configurations of interacting synfire chains and also to random networks. Here the question is whether a random network produces synfire chains spontaneously and perhaps transiently. In due course it should be possible to examine experimentally recorded data (probably from optical measurements) to settle the fundamental question of whether synfire chains occur in real brains.

Logistically, the stay was extremely comfortable: we had a splendid apartment on Goethe Strasse in the University’s Liefmann Haus, with a nice 30 minute walk to Hansa Strasse. Culturally the stay was punctuated by many opportunities to hear good music, especially organ recitals in St. Johann and in the Münster, and also by a number of day trips to Basel to visit some of their museums, as well as a trip to the local beer brewery.

So in summary, besides the various personal activities associated with this research stay at the BCCN, we have begun a new project with many branches that need development in the near future. New and interesting ideas are one of the most rewarding aspects of scientific work; it was a pleasure to have spent time in such an active scientific environment.

George Gerstein, Professor Emeritus
Dept. of Neuroscience, 215 Stemmler Hall
University of Pennsylvania., Philadelphia PA 19104-6074
Personal tools