Klaus Obermayer (Science of Intelligence), “Computational Models of Electric Field Effects and Optimal Control of Neurons and Neural Populations”

Abstract:

The brain is a complex dynamical system with processes operating on different spatial scales. At the macroscopic end one observes global dynamical phenomena, which are called „brain states“ and which are often acompanied by oscillations in different frequency bands or by specific functional connectivity patterns between populations of neuron. A common hypothesis states, that the global dynamics establishes a task-dependent operating point, which is required by individual neurons and local networks to perform information processing tasks. Perturbation experiments are performed, on the one hand, to perform causal analyses into the consequences of this and related hypotheses and, on the other hand, to restore a brain’s operating point in case of dysfunction.

In my talk I will summarize some of our recent modelling work to better understand the interaction between the neural dynamics and external control inputs, taking non-invasive electrical stimulation of neural tissue as an example. I will first present some results on the biophysics of (microscopic) neuron-field interactions and our modelling attempts to propagate these effects to the macroscopic level. In the second part of my presentation I will show, how techniques from Optimal Control Theory can be used to probe controllability aspects of neural systems and to help design efficient ways of steering the neural dynamics.

This talk will take place in person at SCIoI.