Confocal microscopy rejects out of focus light from the object by scanning a pinhole through the object and constructing the image point by point. Volume holographic imaging (VHI) systems with bright-field illumination have been proposed as an alternative to conventional confocal type microscopes. VHI systems are an imaging modality that does not require scanning of a pinhole or a slit and thus provides video rate imaging of 3-dimensional objects. However, due to the wavelength-position degeneracy of the hologram, these systems produce less than optimal optical sectioning because the high selectivity of the volume hologram is not utilized. In this dissertation a generalized method for the design of VHI systems applied to microscopy is developed. Discussion includes the inter-relationships between the dispersive, degenerate, and depth axes of the system. Novel designs to remove the wavelength-position degeneracy and improve optical sectioning in these systems are also considered. Optimization of a fluorescence imaging system and of dual-grating confocal-rainbow designs are investigated. A ray-trace simulation that integrates the hologram diffraction efficiency and imaging results is constructed and an experimental system evaluated to demonstrate the optimization method. This results in an empirical relation between depth resolution and design tolerances. The dispersion and construction tolerances of a confocal-rainbow volume holographic imaging system are defined by the Bragg selectivity of the holograms. It is found that a broad diffraction efficiency profile of the illumination hologram with a narrow imaging hologram profile is an optimal balance between field of view, construction alignment, and depth resolution. The approach in this research is directly applicable towards imaging ovarian cells for the detection of cancer. Modeling methods, illumination design, eliminating the wavelength degeneracy of the hologram, and incorporating florescence imaging capability are emphasized in this dissertation. Results from this research may be used not only for biomedical imaging, but also for the design of volume holographic systems for both imaging and sensor applications in other fields including manufacturing (e.g. pharmaceutical), aerospace (e.g. LIDAR), and the physical sciences (e.g. climate change).

Depuis quelques années, les techniques d'imagerie de fluorescence font l'objet d'une attention particulière, celles-ci permettant d'étudier de manière non invasive un nombre important de processus cellulaires. En particulier, les techniques de fluorescence en rétrodiffusion (FRI pour Fluorescence Reflectance Imaging) présentent plusieurs avantages en termes de facilité de mise en oeuvre, de rapidité et de coût, mais elles sont aussi sujettes à des limites fortes : la pénétration des tissus reste relativement faible (quelques millimètres seulement), et il est impossible d'avoir une information quantitative du fait de la diffusion des photons. L'objectif de cette thèse a été de réduire les effets des signaux parasites afin d'améliorer les performances de la FRI aussi bien au niveau du contraste que de la résolution. Pour ce faire nous avons décidé d'utiliser de nouvelles techniques d'illumination et de détection. Contrairement aux systèmes classiques qui utilisent une illumination et une détection large champ, nous balayons l'objet d'étude avec une ligne laser, des images étant acquises à chaque position de la ligne. On a alors accès à une pile d'images contenant un nombre d'informations bien plus important que dans le cas classique. Trois axes ont été suivis pour l'exploitation de ces informations. Les méthodes développées ont été testées en simulation avec le logiciel NIRFAST et un algorithme de Monte-Carlo mais aussi expérimentalement. Les validations expérimentales ont été réalisées sur fantômes optiques et en in vivo sur petit animal en les comparant à une illumination uniforme plus classique. En améliorant à la fois le contraste et la résolution, ces différentes méthodes nous permettent d'obtenir de l'information exploitable plus loin en profondeur en réduisant les effets néfastes des signaux parasites et de la diffusion. / Intraoperative fluorescence imaging in reflectance geometry is an attractive imaging modality to noninvasively monitor fluorescence-targeted tumors. However, in some situations, this kind of imaging suffers from a lack of depth penetration and a poor resolution due to the diffusive nature of photons in tissue. The objective of the thesis was to tackle these limitations. Rather than using a wide-field illumination like usual systems, the technique developed relies on the scanning of the medium with a laser line illumination and the acquisition of images at each position of excitation. Several detection schemes are proposed to take advantage of the stack of images acquired to enhance the resolution and the contrast of the final image. These detection techniques were tested both in simulation with the NIRFAST software and a Monte-Carlo algorithm and experimentally. The experimental validation was performed on tissue-like phantoms and in vivo with a preliminary testing. The results are compared to those obtained with a classical wide-field illumination. As they enhance both the contrast and the resolution, these methods allow us to image deeper targets by reducing the negative effects of parasite signals and diffusion.