Dr. Fleur Zeldenrust: Assistant Professor at the Department of Neurophysiology, Donders Institute for Brain, Cognition and Behaviour ,Faculty of Science, Radboud University, Nijmegen

Title: Efficient computations in the brain

Abstract: The brain is a unique system, in that it does not only show dynamics, but these dynamics also have a function. It still outperforms any computer we can make, both in energy consumption but also in flexibility. In this presentation, I will explore how the relation between dynamics and function in neural networks can be studied, and how this interaction is determined by structural properties: the connectivity of the network and the properties of each of its nodes. As a model system we use the rodent somatosensory or 'barrel' cortex, the part of the brain that receives input from the whiskers. By viewing the brain as a computer performing algorithms, computational neuroscientists have been capable of deepening our understanding of how complex processes such as perception arise and how the structure of the brain might be optimised to perform these.

Bio: Fleur Zeldenrust, assistant Professor at the Department of Neurophysiology, Donders Institute for Brain, Cognition and Behaviour, Radboud University, Nijmegen, studies how the physical structure of the brain (its ‘hardware’) shapes its information processing and vice versa: how the computations needed for information processing (the ‘software’) are adapted to the physical structure of the hardware. In her ‘Biophysics of Neural Computation’ group, she studies the relation between the physical properties of the brain and its information processing: how are neurons and networks formed so that they can perform functions such as perception? Which characteristics of neurons and networks enhance or limit information transfer? With a training in both Physics and Neuroscience, she has a broad interest in quantitative solutions to all types of scientific problems, but her expertise (PhD at the University of Amsterdam, postdoctoral work at the École Normale Supérieure in Paris) is in the field of Computational Neuroscience.