Skip to main content

September 20, 2011

Speaker: Lennaert van Veen

Title: Meta-bifurcation analysis: glorified number-crunching for dynamical systems analysis

Abstract: In classical dynamical systems analysis, bifurcation theory helps classify the different types of behaviour which occur as we vary model parameters. Although very successful for relatively simple models, this approach has little use for the analysis of physiological models, which typically have many parameters with a large range of admissible values.\r\nI will explain how automated bifurcation analysis of many parameter sets leads to an analysis of robust behaviour, pervasive to the model with all parameter uncertainty considered. As an example, I will present the meta-bifurcation analysis of a mean-field model of human electrocortical activity. This approach unveils non trivial correlations between model parameters and the response to introduction of anesthetic agent, with possible implications of monitoring of anesthesia. With Federico Frascoli, Ingo Bojak and David Liley