|Other Abstract||Our daily life is full of motion information of objects or ourselves, and the formation of motion perception requires collaborations and integrations across multiple sensory modalities. On the one hand, multimodal integration of information helps forming more accurate and reliable percepts. On the other hand, interactions exist between sensory modalities, allowing the perception in one modality to affect the processing of the information from another. The current thesis is about to explore the impact of vestibular self motion information on the visual motion processing.
Study 1 aimed at investigating how the vestibular motion information would influence the processing of visual motion aftereffect. Participants in Experiment 1 were presented with shifting gratings to acquire motion adaptation and were instructed to rotate their head whilst perceiving a motion aftereffect. An additive effect was observed, that is, the motion aftereffect in the opposite direction to the head rotation, compared with that in the same direction, was perceived as shifting faster, which was largely contrary to the observation for real visual motion. These results were basically replicated in Experiment 2 where participants executed the same tasks under two conditions, respectively actively rotating their whole body and being passively rotated by the experimenter thus ruling out the possibility that the additive effect was due to the efference copy in active head rotation. Taken together, the results above suggested that the vestibular motion information had an additive modulation on the processing of motion aftereffect, a specific effect not manifest for real visual motion unless the real visual motion was relatively slow.
Study 2 oriented towards investigating the vestibular role on the head motion reduced flash lag effect and the validity of existing theoretical accounts. Experiment 3 took vestibular and proprioceptive motion information into comparison and suggested the key importance of vestibular contribution in forming the head rotation induced flash lag effect. In Experiment 4, the head rotation was accompanied by an opposite retinal motion. It turned out that the perceptual position of the flash still shifted opposite the head rotation and even outpaced the retinal motion stimulus, an observation that could not be accounted for by existing hypotheses. In Experiment 5, the association of directions between head rotation and retinal motion was manipulated to be perpendicular. After a short term training, the head motion induced flash lag effect was still observed even along the new direction. This finding again suggested that the existing hypotheses could not provide an adequate explanation for the head motion induced flash lag effect; in contrary, it was likely that the head motion induced flash lag effect was also the consequence of the additive modulation of vestibular motion information on visual motion processing.
Study 3 tried to locate the brain areas involved in the head motion induced flash lag effect. Finally, bilateral junctions of inferior temporal gyrus and BA 37 was found to activate while right supramarginal gyrus / BA 40 deactivated during the head motion induced flash lag effect.
Comprehending all findings above, the additive modulation of vestibular motion information on visual motion processing was discovered, and a cross modal bias hypothesis was then proposed as an account. As the neural system develops, the motion signals in the visual and vestibular pathways keeps a relatively fixed relevance with their directions opposite each other. Therefore, the accumulation of sensory experiences keeps the visual and vestibular neurons that represent the corresponding motion directions firing simultaneously and these neurons finally form an association in a Hebbian like manner. Via this association, the vestibular modality is able to add motion signals in the associated (i.e., opposite) direction to the visual pathway and thus modulate the processing as well as bias the perception of visual motion. To avoid interference with the external real motion signal, this bias signal should be rather parsimonious so that it can be overwhelmed by strong forward inputs of visual motion information. But once the forward signal is relatively weak or less than reliable, the cross modal modulating signal would weigh greater against it and thus bias the visual motion signal towards the opposite direction to the vestibular motion signal. This explains the impact of head rotation on the motion aftereffect processing, the perceptual position of the flash in the head-rotation-induced flash-lag effect, and also why similar results were not observed for the strong real visual motion stimuli. This process might involve both visual motion processing and visual-vestibular multimodal integration including sensory conflict processing, relying on the functions of the junction of inferior temporal gyrus and BA 37 (also likely to overlap with MT+) and the supramarginal gyrus / BA 40.|