The representation of ourselves and our environment is based on the combination of information from multiple sensory systems. Each sensory modality is represented within a different frame of reference, in other words each sensory system uses a different code to represent the same properties of the environment (ex. visual stimuli in an eye-centered frame of reference; hand position would be shoulder-centered). Combing this information into a singular coordinate frame is complex. For instance, the eye and shoulder have different centers of rotation, therefore any changes in eye position or body posture will affect the relationship between them. What is still unknown is how the brain integrates these different sources of information into an internal representation, and what effect extra-retinal signals can have on this process. This thesis was designed to investigate the effect of head roll on action and perception. In Experiment 1, we wanted to determine how the integration of vision and proprioception for action was affected by changes in head roll. To investigate this question subjects performed a reaching task at three different head roll positions, where they would experience conflicts between their viewed and actual hand position. In Experiment 2, we examined the influence of head roll on sensory perception. To explore this idea subjects performed a task where they needed to judge the position of their unseen index fingertip relative to a visual target. Our findings reveal that eccentric head roll conditions influence both action and perception – revealed by an increase in movement variability and a decreased ability to discriminate the position of the unseen fingertip relative to visual targets. In summary, we have discovered that introducing eccentric head roll positions affects the perception of ourselves within the environment and the way that we integrate sensory information. / Thesis (Master, Neuroscience Studies) -- Queen's University, 2010-09-21 14:07:09.217

Un simulateur de conduite dynamique est un outil permettant, entre autres, d’étudier les processus d’intégration multi-sensorielle pour la perception du mouvement et la production du comportement de conduite. Néanmoins, les limitations mécaniques des simulateurs, qui imposent des stratégies dynamiques pour simuler le réel, peuvent avoir un impact défavorable sur la perception et le comportement du conducteur. Cette problématique est particulièrement vraie pour la prise de virages qui demeure une situation difficile à reproduire de façon réaliste notamment à cause des variations importantes d’accélérations latérales. Dans ce contexte, cette thèse présente des travaux de recherches permettant de mieux comprendre les processus d’intégration multi-sensorielle (rôle des informations inertielles et visuelles) pour la perception du mouvement en virages et de caractériser l’évolution d’un percept en fonction des conditions de simulation.En conclusion, il a été démontré que la perception des accélérations latérales est basée sur des processus non-linéaires. Le rôle des informations visuelles et inertielles semble donc dépendre des individus et du contexte (notamment du niveau des accélérations latérales) dans lequel ces stimulations sont produites. Sur la base de l’ensemble de ces résultats, des nouvelles pistes d’amélioration du simulateur dynamique SHERPA² de PSA sont proposées. Il est préconisé notamment d’employer un gain du mouvement latéral dégressif avec l’augmentation du niveau d’accélération latérale. / A dynamic driving simulator is a tool, among others, allowing the study of multisensory integration for motion perception and production of driving behavior. Nevertheless, the mechanical limitations of the simulators which impose dynamic strategies to simulate the real can have an adverse negative impact on the driver perception and its behavior. This issue is particularly true for cornering which remains a difficult situation to reproduce in a realistic way, because of massive lateral accelerations variations. In this context, this thesis presents research works allowing to understand the processes of multisensory integration (role of inertial and visual information) for the motion perception in cornering and to characterize the evolution of a percept as a function of simulation conditions.In conclusion, it has been demonstrated that the perception of lateral accelerations is based on non-linear processes. The roles of visual and inertial information seem to depend on the individuals and on the context (notably the level of lateral accelerations) in which these stimulations are produced. Base on of the overall results, new ways for improvement of the dynamic driving simulator SHERPA2 are proposed. It is notably preconized to employ a lateral motion gain digressive with the increase of lateral acceleration.