Michael Breakspear School of Physics and Brain Dynamics Centre, University of Sydney Nonlinear synchronization and desynchronization in neural systems: Evidence and putative function Wednesday 7th, April 14:05-15:55pm, Carslaw Building Room 359. The nonlinear properties of several critical neural processes motivates the application of nonlinear methods to the study of dynamic correlations in brain time series data (EEG, MEG, fMRI). In this talk, the role of complex nonlinear interdependence in large-scale neural dynamics is examined. The theoretical underpinnings of this phenomenon will first be presented, with emphasis given to the phase-space concepts of the 'synchronization manifold' and its 'transverse' stability. In a numerical model of a brain-like neural system, it will be shown that attraction towards the synchronization manifold underlies synchrony between different neural regions, whereas transverse instability permits the itinerant expression of different synchronous configurations. The facility to de-couple and subsequently reconfigure distinct synchronous configurations may be a critical process, ensuring that the brain is able to quickly adapt to changing environmental contexts and generate novel associations. Experimental application of different nonlinear algorithms gives convergent evidence for weak and/or infrequent nonlinear interdependence in scalp EEG and extracranial MEG data recorded from healthy human subjects. An interpretation of the significance of these findings will be offered.