Quantification of diabetic maculopathy by digital imaging of the retina
Diabetic retinopathy is the commonest cause of blindness in the working age population of the United Kingdom, with maculopathy being responsible for most of the visual impairment within this population. In this thesis, the application of digital imaging techniques to the quantification of diabetic maculopathy has been investigated. A suite of programs has been written which quantifies three major aspects of maculopathy, namely microaneurysms, exudates, and macular oedema. The accuracy, repeatability, and reproducibility of these techniques has been investigated in order to assess their potential for clinical use. For each of the three aspects of maculopathy the computer performed at least as well as experienced clinicians when images of comparable spatial resolution were analysed. The problems involved in computerised analysis of maculopathy are identified, and suggestions for improvements are made.