return

Workshop

Informatics in Neuroscience

ZonMw Building, Den Haag, Friday, December 9, 2005



Abstracts

A Scientific Computing Framework for Studying Axon Guidance
Jan Verwer, CWI, UVA, Amsterdam

The proper functioning of the nervous system relies on the formation of correct neuronal connections. During development, neurons project long, thin extensions, called axons, which grow out, over long distances, to form synaptic connections with appropriate target cells. Axons can find their target cells with remarkable precision. Only some of the underlying mechanisms guiding axon growth are known and a major challenge in neuroscience is to understand how all mechanisms involved act in concert to generate the complex patterns of neuronal connections in the nervous system. To address this challenge it seems natural to complement theoretical and experimental neuroscience research with novel computational research based on scientific computing. In this lecture first ideas will be presented for a scientific computing framework directed at supporting axon guidance research by means of mathematical modelling, numerical analysis and computer simulation. The lecture is based on the recent PhD thesis 'On the numerical solution of diffusion systems with localized, gradient-driven, moving sources' of Johannes Krottje of the CWI. This thesis was written within a joint 'Wiskunde Toegepast' project between the CWI, the VU (Arjen van Ooyen) and the NIBR (Jaap van Pelt).

Computational Neuroscience Platforms for Neural Interfacing
Tim C. Pearce - NeuroLab, Centre for Bioengineering, University of Leicester, UK

By exploiting fine-grained parallelism and single clock cycle numerical iteration, spiking neuronal models with realistic synaptic dynamics can be deployed in programmable logic with simulation speeds of at least 5 orders of magnitude faster than real-time. When deployed on commercially available field programmable gate arrays (FPGAs), these physical implementations may then be multiplexed to generate a total of 10^5-10^6 neural elements (dynamical synapses or somas) operating on a single device in biological time - comparable to the number of neurons comprising the nervous systems of Drosophila melanogaster or output stages of the basal ganglia. I will describe the various tricks we used to optimise these low-complexity dynamical neuron designs to achieve exact integration numerical performance. Summed up this platform provides a real-time programmable logic construction kit of commonly used population-based neural model elements, which supports the construction of large and complex dynamical network architectures, ideally suited to coupling with the nervous system. To conclude, I will discuss how such an approach may be adopted for future bidirectional hardware interfacing of neurons.