|Ph.D Student||Zahar Yael|
|Subject||Multisensory Integration: A Physiological Study in the|
Optic Tectum Of the Barn Owl
|Department||Department of Medicine||Supervisor||Professor Yoram Gutfreund|
|Full Thesis text|
To date, the bulk of multisensory (MS) research focuses on discrete, unique events, and also delineated MS integration as manifested through rate code, ignoring the effect of the MS input on the temporal coding.
Initially, I studied responses of MS neurons in the barn owl’s Optic Tectum (OT) to visual, auditory, and bimodal stimuli, specifically focusing on sequences of repeated stimuli. The bimodal stimulus elicited more spikes than the responses to its unimodal components. However, this tendency was history dependent; MS enhancement was mostly apparent in the first stimulus and to a much lesser extent in the subsequent stimuli. In addition, the responses to bimodal sequences were better phase locked to the stimulus than responses to unimodal counterparts.
Previous work showed that the magnitude of the sensory stimulus influences the strength of MS enhancement (the Inverse Effect). I've next studied the history dependence of bimodal enhancement, at different intensity levels. The difference between the enhancements of the first stimulus to later stimuli in the sequence was larger for weak stimuli and less for strong stimuli.
Next I offer a model for the bimodal integration. The model consists of two stages: first, the unimodal streams each undergo adaptation separately. Then the two modalities are integrated based on the coincidence of incoming spikes: the probability of a spike to appear in the output is proportional to its time difference from the closest spike of the other modality. The model qualitatively predicted our empirical results of context-dependent bimodal enhancement, improved phase locking to bimodal stimuli, and intensity-dependent bimodal enhancement.
Most MS studies used simple and synthetic stimuli which are easy to control and analyze, however they do not encompass the full richness in our surroundings. In the last stage of my research I investigated the tectal responses to bimodal combinations of movies and soundtracks. The responses to congruent combination of the realistic stimuli were better synchronized to their temporal structure, compared to unimodal or incongruent bimodal stimuli. Next I've used an algorithm developed to detect salient events in movies (Barzely et al., 2007). The algorithm successfully predicted the occurrences of enhanced MS responses at certain salient events in my stimulus.
To summarize, I have characterized two coding strategies implemented by tectal neurons: an improved temporal coding (better phase locking) for bimodal stimuli compared with unimodal, and bimodal enhancement that is context-dependent.