Recognition of dynamic patterns during auditory scene analysis. (360G-Wellcome-099750_Z_12_Z)

To make sense of complex auditory input signals, the mammalian auditory systemmust group the elements of those signals together into identifiable auditory 'objects' or 'streams'. Neither the perceptual rules that govern this groupingnor the neuronal mechanisms by which it is achieved are fully understood. Thisproject proposes to investigate the role of stimulus directionality in sound grouping. We hypothesise that the auditory system is able to identify the directionality of sounds (specifically, increasing or decreasing pitch, level or spatial azimuth) and use this as a cue to group together sounds with similar direction. This hypothesis will be tested through psychoacoustic experiments. We also propose to investigate the neuronal mechanisms underlyinggrouping by directionality. We hypothesise that neuronal populations in the auditory system explicitly represent the directionality of stimuli, and will record neuronal responses in the ferret auditory system to test this. We will investigate whether this is a predictive code, where future values of the stimulus are predicted by the direction of change of past values and neuronal responses are modulated by the deviation of the stimulus from the predicted value. Using parallel neuronal recordings, we will also investigate whether this representation provides a possible basis for the perceptual grouping effects observed psychoacoustically.