In this thesis, a phase sensitive interferometer is successfully implemented to perform birefringent object surface-profile measurement, based on a polarisation adjustment approach. Using monochromatic light, a novel polarization interferometric method is developed, incorporating the birefringence technique and a waveplate. In our experiments, a birefringent wedge is designed for generating carrier fringes in the polariscope. Retardance is calculated from phase shifting using a phase matching technique. The accuracy of the method has been demonstrated to have an error of less than 0.02 radians. The accuracy and resolution quantitative analysis presented in this thesis can be used to determine accurately the phase-shifting interferometry for high-precision surface profile and bio-structure, such as fibre and collagen measurements with low cost. FFT technique and phase-stepping methods are described to determine birefringence within the cornea. The distribution of human corneal lamellar collagen is determined through a microscopic technique using the combination of a circular polariser and a quarter-wave retarder. A quantitative measure of corneal birefringence is achieved by phase unwrapping. The experimental findings of elliptic and hyperbolic populations of collagen fibrils may explain the optical phenomena of central corneal retardation with biaxial-like behaviour in more peripheral areas. A low cost, simple, and direct approach has been developed to make the required microscopic measurement. The traditional transmission system is improved by applying a reflection system with an LED light source and is suitable for the analysis of the birefringent cornea structures in vivo. A further instrument based upon a synthetic aperture approach has been created with the purpose of measuring the three dimensional birefringence structure of the cornea. The concept of the instrument is a combination of the parallax between individual lenses and the numerically generated planes of focus to visualise the phase structure.

The knee is a complex joint, prone to instability and damage, meaning a complicated architecture of soft tissues is necessary to ensure any stability of the joint. These structures are innervated, and play an important role in both proprioception, the sensing of a body’s own limb positions, and nociception, the sensing of painful stimuli. The purpose of this project has been to develop a computational model that can replicate the behaviour of the mechanical sensing nerve endings in the knee joint. An adapted Hodgkin-Huxley model has been developed and used to simulate the behaviour of the nerve endings. These models have been coupled with a three dimensional finite element model of a feline knee joint, which has been built with use of x-ray CT and MRI scans of a cat’s hind limb, allowing neural responses to be predicted as the position of the knee joint changes. Once the behaviour of the complete model has been verified, through comparisons with recordings of neural responses in the literature, it was possible to observe the effect of removing a soft tissue structure on the neural response. The anterior cruciate ligament (ACL) was removed from the model, and a series of tests run to determine the effect of ligament damage on neural response. It was predicted that removing the ACL from the knee joint can increase the neural responses to changes in knee position, agreeing with data in the literature. This could indicate an increase in pain at the joint, and could help with understanding the causes of pain and changes proprioception experienced by patients with damaged ACL.