Thesis defense Ana Todorovic (Donders Series 184)

29 June 2015

Promotor: Prof.dr. P. Hagoort, copromotor: dr. F.P. de Lange

Predictive adaptation in early auditory processing

Neural adaptation is a ubiquitous property of sensory processing. The response to a single stimulus varies with the neural context into which that stimulus arrives when it meets the sensory system, and this context is shaped by the history of stimulation. The sensitivity of the sensory cortex to an incoming stimulus is largely modulated by whether the same stimulus was recently observed, but it is also dependent on more sophisticated properties of sensory processing, such as the distribution of previous stimulus events, local transitional rules between stimuli, their temporal characteristics, global regularities, as well as endogenous states such as recognition of an object, expectation and attention. In this thesis, I focus on local statistical regularities in transitions between auditory tones, and attempt to recast neural adaptation as a predictive phenomenon: the more surprising, statistically speaking, a stimulus event is, the more neural activity it will elicit. This research demonstrates that the sensory brain processes stimulus regularities even when unattended, and adapts its neural response to them.