PriorDynamics · The neural dynamics of perceptual priors in audition

To maintain a coherent and continuous percept over time, the brain relies on past sensory information to predict forthcoming stimuli. Combining novel behavioural methods with advanced neuroimaging techniques, the present research project aims to determine the neural mechanisms by which the auditory system uses information from the recent past to overcome signal noise and ambiguity. Specifically, we test the hypothesis that sensory predictions involve neural oscillations at alpha rhythm, ~9-10 Hz, that mediate the propagation of perceptual priors. The first part of this research project investigates the neural structures underlying such predictive oscillatory mechanism. To this end, we adapt a novel time-resolved sampling technique used to examine rhythmic fluctuations in perceptual performance for functional magnetic resonance imaging (fMRI). We hypothesise that cortical and subcortical activation, especially in the auditory cortex and thalamus, will exhibit rhythmic modulations that correlate with oscillations in decision bias during the detection of a brief auditory signal masked by white noise. The second part of this project examines whether the resolution of perceptual ambiguities by prior contextual information involves similar oscillatory mechanisms. For this purpose, we combine the time-resolved sampling technique with a classic paradigm (involving Shepard tones) for inducing ambiguous pitch shifts. By presenting a single tone before the ambiguous stimulus, listeners can be biased toward a perceived upward or downward pitch shift. If resolving perceptual ambiguities involves oscillatory mechanisms, we expect to observe periodic fluctuations of decision bias (that is, the tendency to make certain responses) at alpha rhythm over time. The results of this research project will shed light on a potentially crucial and yet unknown core process of perception, useful in every interaction we have with the world.