Visual conditions such as retinitis pigmentosa and Usher syndrome can gradually cause tunnel vision. Patients with these conditions usually face difficulties with navigation, avoiding obstacles, and performing visual search. Loss of mobility can affect patients' independence and quality of life. One of the rehabilitation strategies for patients with tunnel vision is the use of optical aids to enhance mobility performance. The main method used to evaluate the usefulness of optical aids is the patient’s subjective report after extended wear. In order to evaluate optical aid effectiveness in the clinic, a new test based on the visual search paradigm was designed to assess the patient's visual awareness. This was named the assessment of visual awareness (AVA) test. The main aim of this study was to develop the AVA test, establish its sensitivity, validity and repeatability, and then use it to investigate the efficacy of optical aids in this group of people. The AVA test consists of 32 peripheral targets presented at four different locations: 1st annulus (at 5° from the central fixation), 2nd annulus (10°), 3rd annulus (20°) and 4th annulus (30°)). In this study, the peripheral targets were presented singly against a spatial noise background in a presentation area of 81° H × 62° V. Participants were allowed to use head and eye movements and were asked to search for and locate each target. The detection time (DT) was recorded. A new, sensitive and easy to set up indoor mobility course was also designed and validated prior to its use in validating the AVA test. A total of 50 normally sighted participants with simulated tunnel vision (TV) (5° to 20°, in 5° steps) and 20 patients with TV (retained field 4° to 21°) were tested. The AVA test was found to be responsive to the change in field of view (FoV) and to the target locations in both groups of participants. In the simulated group, a significant relationship was found between FoV and DT at each annulus (r ranging from -0.55 to -0.77, p < 0.0001). A significant relationship was found between target location and DT within each FoV size (20°, 15°, 10° and 5°) (r ranging from 0.53 to 0.84, p < 0.0001). In the TV patients, a statistically significant relationship was found between FoV and DT at each annulus (r range from -0.40 to -0.60, p < 0.05). The target location was shown to have a significant relationship with the DT within each FoV size (r ranging from 0.50 to 0.60, p < 0.05). Finally, the AVA test was found to be significantly related to the simulated TV participants' performance on the indoor mobility course. The AVA test was used to assess the efficacy of three optical aids: the partial aperture prism (10 patients), the Tri-field prism (10 patients) and the reverse telescope (4 patients). The AVA test showed no significant improvement in DT with either of the prisms and the participants did not find these aids helpful. DT with the reverse telescope improved, but none of the participants were willing to use these on extended trial. The AVA test gave clear indications of the efficacy of each aid, a result which could affirm the importance of the AVA test. In conclusion, the AVA test was found to be sensitive, valid and repeatable. DT did not improve in either of the optical aids which were found to be unsuccessful, suggesting that the AVA could be a promising clinical test. However the aids which showed improved DT were not evaluated over the longer term, and therefore did not allow full evaluation of the AVA test.

La perte de vision périphérique (vision tubulaire) a pour conséquence de nombreux déficits qui réduisent fortement l'autonomie des personnes qui en sont atteintes et par conséquent leur qualité de vie. Nombre de ces déficits témoignent d'une cognition spatiale dégradée mettant en jeu des relations étroites entre vision périphérique et représentations spatiales. Le double objectif de ce travail de thèse consiste à mieux comprendre la nature de ces relations et, sur la base de ces connaissances, contribuer à l'émergence de dispositifs d'assistance plus adaptés et performants, capables de suppléer l'absence de vision périphérique. Dans un premier temps, je me suis intéressé au rôle de la vision périphérique dans la construction de représentations spatiales égocentrées. J'ai d'abord collaboré à une expérience de psychophysique impliquant la détection en vision périphérique de cibles visuelles situées soit droit-devant, soit excentrées par rapport à l'axe du corps. En mesurant les temps de réaction nous avons pu démontrer que les sujets humains répondaient plus rapidement aux cibles présentées droit-devant qu'aux cibles excentrées. J'ai pris en charge une deuxième étude complémentaire portant sur le lien entre ce traitement sensoriel privilégié du droit-devant en vision périphérique et la dynamique des saccades oculaires de recentrage. En comparant les dynamiques de pro-saccades et d'anti-saccades de recentrage ou d' " excentrage ", j'ai pu mettre en évidence que la supériorité dynamique des saccades de recentrage guidées par la vision périphérique reposait à la fois sur des facteurs sensoriels et oculomoteurs. Ces travaux révèlent l'intégration précoce de signaux visuels et oculomoteurs en vision périphérique, pouvant servir à localiser les éléments visuels par rapport à soi et à privilégier le traitement des éléments situés dans l'axe droit-devant. Dans un deuxième temps, mes travaux ont abordé l'influence de la vision périphérique dans les représentations spatiales allocentrées. Pour étudier l'implication de la vision périphérique dans le codage allocentré, j'ai réalisé une expérience impliquant une tâche de pointage vers des cibles visuelles en vision centrale, accompagnées ou non d'indices visuels à différentes distances en vision périphérique. Les résultats obtenus montrent que des indices visuels capturés par la vision périphérique peuvent effectivement contribuer au codage allocentré d'une cible fixée, et ce même lorsque ces indices périphériques doivent être extraits de scène visuelles complexes en moins de 200 ms. Dans une étude complémentaire, j'ai montré que l'utilisation de ces indices allocentrés situés en périphérie avait un véritable rôle fonctionnel, accélérant les recherches visuelles. Ainsi, ces travaux révèlent une implication fonctionnelle forte de la vision périphérique dans l'extraction des relations spatiales entre éléments présents dans l'environnement visuel. J'ai voulu, durant la dernière partie de ma thèse, initier le développement d'un dispositif d'assistance dont la finalité est de suppléer les fonctions spatiales, égocentrées et allocentrées, de la vision périphérique. Comme preuve de concept, mon travail a consisté à concevoir et développer un dispositif tactile placé sur le poignet et capable de communiquer la position spatiale d'objets d'intérêt pour en simplifier leurs recherches visuelles. Les résultats obtenus chez des sujets sains avec un champ visuel artificiellement réduit (10°) montrent que l'interface tactile permet d'accélérer par trois la vitesse de recherche visuelle. Des résultats similaires ont été observés chez une personne atteinte de glaucome (champ visuel de 10x15°). Ma thèse pluridisciplinaire permet d'apporter un nouvel éclairage sur l'implication de la vision périphérique dans la construction de représentations spatiales, et elle propose de nouvelles pistes pour le développement de dispositifs d'assistance adaptés aux personnes atteintes de vision tubulaire. / The loss of peripheral vision (tunnel vision) leads to numerous deficits, reducing both independence and quality of life. These deficits reflect spatial cognition impairments, and highlight the close relationship between peripheral vision and spatial representations. This thesis has two main objectives: reaching a better understanding of the nature of these relationships, and using the acquired knowledge in order to propose adaptive, performant and innovative assistive devices able to overcome the peripheral loss. At first, I address the role of peripheral vision in egocentric space coding. I collaborated in a psychophysics experiment, involving detection of visual objects placed in peripheral vision. The visual objects formed similar images on the retina and differed only with respect to their egocentric location: either straight-ahead or eccentric with respect to the head/body midline. We found that straight-ahead objects elicit consistently shorter behavioral responses than eccentric objects. I took in charge a second study evaluating the link between the privileged sensory processing of the straight ahead direction and the dynamic of ocular saccades. Comparison between centripetal and centrifugal pro-saccades and anti-saccades revealed that the superior dynamic of centripetal saccades comes from both sensory and oculomotor factors. These works reveal the early integration of both visual and oculomotor signals in peripheral vision, leading to egocentric representations in which the straight ahead direction is highlighted. Secondly, I investigated the influence of peripheral vision in extracting allocentric spatial representations. In order to assess the role of peripheral vision in allocentric coding, I performed a memory-based pointing task toward previously gazed targets, which were briefly superimposed with visual cues placed at different eccentricities. The results showed that visual cues in peripheral (>10°) vision can contribute to the allocentric coding of a fixated target. A complementary experiment showed that these peripheral allocentric cues play a functional role, notably by facilitating visual searches. These works highlight the importance of peripheral vision in extracting functional spatial relationships between distant elements of the visual environment. Finally, I wanted to promote the development of new assistive devices, able to substitute both egocentric and allocentric spatial functions of the peripheral vision. As a proof of concept, I designed and evaluated a tactile interface mounted on wrist, communicating the spatial location of specific objects and facilitating visual search. Results showed that healthy subjects with artificial tunnel vision (10°) were able to increase by three visual search speeds thank to this tactile interface. Similar results were obtained on a glaucoma subject (field of view 10x15°). My multidisciplinary thesis highlights new roles of peripheral vision in spatial representations and proposes an innovative solution to develop assistive device for tunnel vision.