The objective of the research is to provide the clinician with a simple system for making quantitative measurements that is comparable to the services of a skilled orthoptist. As many optometrists do not possess the necessary equipment for making such assessments, this technology would significantly enhance their referral capabilities. Reviews of the tests involved and instruments available are detailed in Chapter II and III. Chapter IV presents the various validation experiments carried out on the Tobii X120 eye tracker, concentrating on technical specifications such as, linearity range, optimum measurement distance and setup (with and without chin rest). We also investigated the effect of wearing different types of lens materials and pupil size measurements on the eye tracker system. We established the inter-examiner agreement of cover test measurements on groups of non-strabismic and strabismic subjects. This study, as detailed in Chapter V, involved collaboration with two clinical orthoptists. We found a good inter-examiner agreement for both the non-strabismic and strabismic cover tests. We further investigated the use of the eye tracker in providing more reliable findings for cover test measurements as compared to the conventional cover test (Chapter VI). Finally we extended the investigation to a number of different clinical subjects attending the Bristol Eye Hospital (Chapter VII) in order to evaluate our purpose-developed monocular calibration routine. Performing quantitative eye movement analysis will provide valuable additional information in any clinical investigation of patients with ophthalmological and/or neurological disorders, leading to greater precision in diagnosis. Traditional methods for the evaluation of oculomotor disorders rely on the diagnostic and therapeutic judgements by the examining clinicians and subjective responses from the patient. However, the use of currently available eye movement recording system will provide valuable alternatives for obtaining more objective and quantitative measurements.

The centrifugal visual system (CVS) is a feedback pathway of predominantly visual information from the brain to both eyes, but principally to the contralateral retina. The CVS is often considered to be something of a peculiarity, regarded as being specific to birds (Aves). Indeed, so-called ‘higher’ vertebrate species are assumed not to even possess such a centrifugal pathway when, in fact, an efferent projection to the retina has been conclusively demonstrated in all vertebrate groups (including humans). Perhaps this point of view reflects the lack of progress made in the elucidation of function in the bird, the dominant model for CVS research in the 120 years since being first described. In the series of experiments presented here, I have begun to investigate the role of the CVS in the modulation of eye growth. In addition, I have addressed a number of unknowns that exist regarding the midbrain connectivity of the CVS. In a series of four parallel lesion experiments, the centrifugal efferent pathway to the retina was unilaterally disrupted in post hatch chicks, raised under different developmental conditions. Under normal visual conditions but in the absence of centrifugal efferents, eyes contralateral to the lesion developed shorter eyes and moderate, relative hyperopia (long-sightedness). In contrast, under constant light conditions, ipsilateral eyes became significantly shorter than fellow (i.e. contralateral) eyes. Compensation for, and recovery from, plus and minus lens-imposed defocus in contralateral eyes was largely unaffected. Centrifugal efferents emanate from two distinct midbrain populations: the isthmo-optic nucleus (ION) and the surrounding scattered cells within the ectopic area (EA). From experiments using pathway tracing techniques, I have demonstrated that, unlike cells of the ION, EA cells do not receive input from primary visual areas. In addition I present evidence for a possible ‘cross-talk’ pathway between centrifugal cells on either side of the midbrain, and discuss its potential involvement in the normally symmetrical eye growth.

The studies described in this thesis aim to assess the importance of normal colour vision in visually demanding, colour-related tasks that are often safety-critical and aim to improve our understanding of how congenital deficiencies can affect the processing of colour signals and the corresponding changes in visual performance. The first study compares the colour vision requirements within different professional environments. 519 subjects were tested: 141 normal trichromats, 268 deutans and 110 protans. All subjects carried out the Ishihara 38-plate test, the CAD test and the Nagel anomaloscope, and sub-populations were examined with the AO-HRR plates, the Farnsworth D15, the City University test (2nd Ed.), the Holmes-Wright type A and B lanterns. Inconsistencies of outcome amongst the various tests and potential alternative practices are discussed. The second study focuses on understanding the discrepancies in performance observed on lantern tests when the subject’s task is to report the colour of small signal lights presented against a dark background field. These conditions were simulated using a psychophysical luminance pedestal technique. Variations in the measurement of chromatic sensitivity over the visual field, as well as the detection of targets where colours are combined with luminance contrasts, are discussed and explanations considered with regard to underlying retinal physiology. The last study investigates the use of colour signals in ATC (air traffic control) applications. The work carried out addresses current failings in acceptance criteria for applicants, and provides alternative methods of assessing suitability. The chromatic discrimination thresholds of normal trichromats and colour deficient subjects were related to performance on a set of visual search tasks selected to be more representative of typical colour usage in large field visual displays. Display parameters under which the performance of colour deficient observers could be comparable to that of normal trichromats are examined with regard to updating occupational acceptance criteria.

Conventional imaging devices only capture a part of the total information carried by the light. A new generation of imaging devices, plenoptic systems, use an array of micro lenses to codify the light coming from an object into a four dimensional function called the light field. The final image is then obtained after post processing computations on the light field. In this work plenoptic imaging devices are analysed using a wave optics approach. A platform to simulate light propagating under the Fresnel approximation in a generic optical system was developed in MATLAB. An optical system can be modelled as the composition of two basic operators: the free space propagation and the lens. The first one was implemented developing an original method derived from the angular spectrum of plane waves theory of propagation. The second was implemented using a phase mask. The code was developed to optimize the signal to noise ratio and the computational time. Two different configurations of plenoptic imaging systems were simulated. The first is the plenoptic 1.0 configuration. The general theory of plenoptic 1.0 and the post processing algorithms presented in the literature were verified using the simulation platform. The effects of diffraction were also evaluated and an original refocusing method is presented. For the second configuration, plenoptic 2.0, a full study of the optical resolution has been made and a detailed analysis of the effects of diffraction is presented. The results achieved with the simulations have been used to design a working prototype of a plenoptic microscope. This novel wave optics approach enables us to quantify for the first time in the literature the effects of diffraction on this class of devices. In plenoptic 1.0 diffraction is a source of noise due to the crosstalk between neighbouring lenslets. In plenoptic 2.0 systems the optical resolution is directly proportional to the magnification of the lenslet array. A small magnification leads to a high directional sampling but at the same time to a loss of optical resolution. The finite dimensions of the lenslets together with the wave nature of light produce a physical limit to the amount of information that can be achieved sampling the optical fields with those kind of devices.

Static automated perimetry of the central 30˚ is the most often used visual field test in glaucoma patients. Short test durations are achieved by focusing on a central region, which constitutes ~20% of the visual field. However, ignoring the periphery may sacrifice information on how patients are affected functionally. Peripheral vision is important for guiding attention, balance and mobility. An efficient standard automated examination for the peripheral visual field has not been established yet. This thesis aims to lay groundwork for the development of such a test. I introduce a kinetic automated test, which estimates an isopter with three repeated presentations per meridian. I ask whether measuring a peripheral isopter adds information to central visual field test results, investigate retest reliability and evaluate the efficiency of test procedures with repeated presentations through computer simulations. Moreover, I investigate how visual field thresholds obtained with static and kinetic stimuli relate to each other and examine the influence of stimulus sizes III and V on static threshold estimates. I also investigate the relationship between response variability and contrast sensitivity in the peripheral visual field. Based on the results, I suggest using repeated presentations in automated kinetic tests. I demonstrate that data driven computer simulations are useful for the development of efficient automated kinetic perimetry. The frequency-of-seeing results suggest that response variability to static stimuli in the far periphery is lower than suggested by previous data (Henson et al., 2000). This is relevant to future computer simulations of peripheral visual field tests with static automated perimetry. As a future avenue for examining the visual field periphery I propose a combined static kinetic automated visual field test, which combines a peripheral isopter as a region of interest with static stimuli inside this region. In a separate investigation, I examine the influence of visual field damage on reading performance and evaluate the relationship between reading performance and eye movements, using a within-patient between-eye study design in glaucoma patients with asymmetrical visual field loss. Between-eye reading performance was affected by visual field loss and co-occurred with specific eye movement patterns. The within-patient between-eye design appeared to be useful for investigating the relationship between visual field loss and functional disability.

Dry age-related macular degeneration (AMD) is a common eye condition, which causes progressive loss of central vision; there is currently no treatment. The live studies presented in this thesis aimed to explore different aspects of everyday visual function and the experience of living with dry AMD. In the first study, a systematic review of the literature surrounding quality of life and "real-world" visual ability in AMD was conducted. Several domains of everyday life affected by AMD were identified; these had been investigated using a wide variety of study designs. However, only 4% of studies meeting inclusion criteria specifically investigated individuals with dry AMD. In the second and third studies, people (>60 years, logMAR binocular visual acuity of 0.7 or better) categorised with varying severities of dry AMD performed two previously validated computer-based real-world visual tasks. Comparisons for both tasks were made against a 90% normative limit for the outcome measures established in age-related peers with healthy vision. In a search task, participants were instructed to find items within digital photographs of everyday indoor and outdoor scenes. Sixty-one percent of participants with dry AMD, (including all of those with late dry AMD) exceeded the 90% normative limits for average search time; this was statistically significant (Fisher's exact test, p < 0.0001). In a face recognition task, participants completed a modified version of the Cambridge Face Memory Test. Percentage of correctly identified faces was used as an outcome measure. Five (17%) participants with dry AMD scored worse than the 90% limit (Fisher's exact test, p=0.46); four of these had fovea-involving geographic atrophy. In the fourth study, volunteers with dry AMD described their visual symptoms in an interview and were asked to comment on the realism of a photograph typically used to simulate vision in AMD. Interview transcripts underwent content analysis. The most frequently used visual symptom was blur (n=13) followed by missing part/s (n=10) and distortion (n=7). Seventy-six percent (95% confidence interval 53-92%) of the participants rejected the realism of a popular image often used to portray the visual symptoms of AMD. In a fifth study, a novel test measuring mobility anxiety was introduced. Participants were shown a series of short movies of navigation through real-world mobility scenarios and were asked to press on a button during scenes which would cause them anxiety or discomfort. Pressure on the button was recorded throughout the test and this was used as an outcome measure. People with intermediate and late AMD applied more pressure to the button, on average, than those with early and no AMD (Kruskal-Wallis test, p=0.04). Situations involving negotiating stairs were identified as particularly problematic. To conclude, these studies spotlight the problems people with dry AMD have with visual tasks. In addition, the tests described in this thesis may have potential to be used as patient- based outcome measures for clinical trials for future treatments for dry AMD. Moreover, findings reported in this thesis may help clinicians with patient management and expectations, and should inform future patient, public and professional education about dry AMD.

Femtosecond lasers were introduced in ophthalmology initially for flap creation for LASIK. This thesis describes a body of work undertaken by the author exploring the possibility of using these lasers in corneal and refractive surgery. The use of the femtosecond laser in corneal and refractive surgery offered the prospect of better precision with respect to their accuracy in depth cut, and the smoothest of lamellar interface. The development of multiple laser platforms allowed us to perform comparative studies in both ex vivo/animal and clinical studies and to explore the prospect of a new refractive procedure, lenticule extraction and also lenticule re-implantation. The laser proved to be accurate in its vertical depth cutting and following optimization was able to cut a smoothe lamellar interface. The clinical study showed the laser to be safe and effective. The comparative studies showed the superiority of the lower energy femtosecond laser on IOP rise, without compromising on clinical outcomes, which were the same for both lasers. Femtosecond laser lenticule creation was optimised in animal models and then shown to be safe and efficacious in a clinical study. The wound healing benefits of an 'all in one' femtosecond laser procedure were evident, in both animal as well clinical studies. Lenticule reimplantation was shown to be effective in both the rabbit and monkey models. The use of the femtosecond laser is set to increase in ophthalmology. The work in this thesis has provided fundmental ex vivo, animal and clinical benefits on the the use of femtosecond lasers in corneal and refractive surgery. It has also envisioned a concept of lenticule re-implantation for future clinical use.

Age-related macular degeneration (AMD) is a major cause of irreversible central sight loss in the elderly. Many factors affect disease onset and progression, and these include age, environmental stressors such as smoking, diet, inflammation and genetic polymorphisms, all of which are likely to influence metabolism. In some complex diseases, metabolomics, which involves the identification of a metabolic fingerprint in a biofluid or tissue, has been shown to discriminate metabolic changes associated with different disease processes and to identify specific phenotypes. We have applied metabolomics, in the first study of its kind, to analyse, using NMR spectroscopy, both serum and urine metabolic characteristics in patients with dry and wet AMD (n=104). NMR spectral analysis showed good clustering as well as separation among the serum and urine from dry and wet AMD patients. The results show that metabolite profiles can distinguish dry and wet AMD, but that the pathways involved, glycolysis, urea cycle and Kreb’s cycle, are involved in both forms of the disease. It is likely that the pathogenesis of dry and wet AMD is similar and that the severity of ocular damage and systemic inflammation would account for the distinguishing profiles. These data support the use of metabolomics in identifying biological pathways involved in pathogenesis of AMD, but not in the diagnosis or prognosis of disease.

Age-related macular degeneration (AMD) is of increasing concern given the ageing population, and the associated economic and social burdens. Vision-related quality-of-life (QoL) is arguably one of the most important factors in the management of those with AMD. Consequently, there is a clear need for an understanding of the clinical outcomes that influence vision-related QoL in order to inform management strategies. The principle aim of the studies described herein was to determine the factors that predict vision-related QoL in those with AMD, over 1 year. Experimental procedures were undertaken at baseline (n=52 individuals with AMD) and repeated after 1 year (n=32 individuals with AMD). These included: visual acuity, contrast sensitivity, reading speed, microperimetry, optical coherence tomography and fundus photography. A questionnaire interview included assessment of vision-related QoL (Impact of Visual Impairment questionnaire), health status (EQ-5D), level of depressive symptoms (PHQ-9) and well-being (Warwick-Edinburgh Well-Being Scale). At baseline, the optimum multiple regression model accounted for 41% of the variance in vision-related QoL and included Mean Total Deviation or Mean Sensitivity with level of depressive symptoms. After 1 year, the optimum model to predict change in vision-related QoL accounted for 43% of the variance and included baseline contrast sensitivity and change in health status and reading speed. The most clinically useful measures of visual function, in identifying those with a reduced QoL or those at risk of a reduced QoL were contrast sensitivity, microperimetry, and reading speed. These outcomes may allow a better understanding of vision-related QoL if they were adopted in a clinical setting. In conclusion, the studies provide sufficient evidence to encourage a review of the clinical outcome measures most relevant to vision-related QoL.

The studies comprising this thesis were conducted to examine the potential of a range of cross-linking therapies; in particular, to investigate the effect of a novel cross-linking therapy (involving a bacteriochlorophyll derivative and near-infra red illumination (WST-D/NIR)) on the structure of the cornea, and to develop a trans-epithelial riboflavin/Ultraviolet-A (UVA) corneal cross-linking protocol that was equally effective to that of the standard protocol (SCXL), without the need for epithelium removal. A number of laboratory techniques were used to investigate changes in the structure of the cornea and its biochemical and biomechanical properties following cross-linking. X-ray scattering and electron microscopy data provided evidence that treatment with WST-D/NIR resulted in no change in corneal collagen organisation and confirmed its potential as an alternative to riboflavin/UVA cross-linking for stiffening diseased or surgically weakened corneas. Enzyme digestion studies and strip extensometry were performed to compare the effectiveness of newly developed riboflavin/UVA protocols to that of the SCXL protocol in terms of their ability to increase the enzymatic resistance and stiffness of the cornea. The studies indicated that the intensity and distribution of cross-links formed within the cornea vary with different protocols, and that the outcome of trans-epithelial riboflavin/UVA cross-linking may be significantly enhanced through the use of higher concentrations of riboflavin, a longer duration of iontophoresis and the use of pulsed and higher energy dose UVA. Although the precise amount of CXL required to prevent the disease progression is still unidentified, the full stromal depth of CXL post SCXL treatment might not be needed, therefore, modified trans-epithelial protocols identified in this thesis may be sufficient to prevent disease progression. Further clinical studies, especially randomized prospective trials, are, however, required to confirm the encouraging results of these modified procedures.