Image segmentation and analysis methods applied to anterior eye optical coherence tomography

Abstract

This thesis aims to expand the current knowledge in the field of AS-OCT image analysis. To achieve this, a range of studies were designed in order to characterise, understand and explore how various programming paradigms have been applied to AS-OCT datasets. Different classification and semantic methods were developed to understand the effect of diverse quality datasets, and the potential clinical applicability. These studies showed different programming paradigms and architectures, which were tested to see performance. The segmentation of OCT images with rigid contact lenses was also considered as potential clinical application of these methods.