|Ph.D Student||Reches Amit|
|Subject||Saliency Coding in the Gaze Control System of the Barn|
Owl: Unimodal and Bimodal Aspects
|Department||Department of Medicine||Supervisor||Professor Yoram Gutfreund|
|Full Thesis text|
The brain is constantly provided with an immense amount of sensory information of which only a small fraction can be further processed. Determining the significance of a stimulus is termed saliency coding, where sensory information is rapidly integrated across space and time: Change detection is an example of temporal saliency coding; Multisensory integration has a strong effect on spatial saliency. Saliency coding is intimately linked to gaze control. In this work, I studied different aspects of change detection and multisensory integration in the gaze control system of the barn owl.
Stimulus specific adaptation (SSA) has been recognized as a neural correlate of change detection. Initially, I characterized unimodal SSA in the localization pathways of the barn owl. In both the Optic Tectum (OT) and the Arcopallium, the two final sensory stations of midbrain and forebrain auditory localization pathways respectively, neurons responded stronger to rarely presented sounds than to the same sounds when presented frequently. Moreover, this history dependent phenomenon occurred for a wide variety of acoustic parameters, demonstrating that these regions were sensitive to the abstract feature of change itself. Neural responses from the ICx, the main auditory input to the OT, did not show similar SSA effects for all attributes, pointing to a top-down effect.
I next investigated the tectofugal pathway, a visual pathway involved in saliency coding. Neurons in the nucleus rotundus and the Entopallium (E) responded to auditory as well as to visual stimulation. Inactivation experiments suggested that this auditory signal originates in the OT. Responses to visual stimulation in the E were strongly enhanced when presented in spatial congruency with simultaneous auditory stimulation (multisensory enhancement). These results are in line with the notion that a visual saliency coding pathway should benefit from the appearance of a spatially congruent auditory signal.
Finally, I studied the interaction between change detection and multisensory enhancement. Entopallial neurons demonstrated sensitivity to rare presentations of visual, but not auditory stimulation compared to the same stimuli when presented frequently. However, when auditory and visual stimulation were presented together, the ability to detect change dramatically increased. Additionally, bimodal enhancement was context dependent - enhancement was robust for deviants, but strongly reduced for frequent appearances. These effects were present for congruent, but not incongruent, bimodal stimulation. To summarize, these results suggest that the different neural substrates of the gaze control pathways work in concert to encode salient events in space and time within a multisensory environment.