Modelling of light propagation in retinal tissue

ABSTRACT Techniques used in retinal imaging provide a unique method for gaining information about biological structures and processes without being invasive. Applications within the fields of medicine and clinical diagnosis provide great scope for such research. Being able to measure haemoglobin oxygenation from retinal images provides useful information in the diagnosis of conditions such as diabetic retinopathy, age-related macular degeneration and glaucoma. We describe how existing methods have been used to gain information from retinal images. Problems of calibration and difficulties encountered in validating the various models are also discussed. Existing techniques to model multi-layered tissue, such as Monte Carlo methods and radiative transfer approaches, are explained and their respective advantages and disadvantages are highlighted. A proposal to employ a standard fundus camera, adapted to accommodate a liquid crystal tunable filter, is presented and the characteristics required of the images are outlined. We finish with a discussion of the techniques deemed to be the most promising and how the captured images can be used to validate them