Quantitative methods in ocular fundus imaging: Analysis of retinal microvasculature

Several diseases including diabetes, hypertension and glaucoma are known to cause alterations in the human retina that can be visualized non-invasively and in vivo using well established techniques of fundus photography. Since the treatment of these diseases can be significantly improved with early detection, methods for the quantitative analysis of fundus imaging have been the subject of extensive studies. Following major advances in image processing and machine learning during the last decade, a remarkable progress is being made towards developing automated quantitative methods to identify image-based bio-markers of different pathologies. In this paper, we focus especially on the automated analysis of alterations of retinal microvasculature - a class of structural alterations that is particularly important for early detection of cardiovascular and neurological diseases.