Colour Texture Segmentation of Tear Film Lipid Layer Images

摘要

Dry eye is a symptomatic disease which can be diagnosed by several clinical tests. One of them is the evaluation of the interference lipid pattern and its classification into one of the Guillon categories. Previous researches have automatised this manual test, saving time for experts and providing unbiased results. However, the heterogeneity of the tear film lipid layer makes its classification into a single category per eye impossible. For this reason, this paper presents a first approximation to segment tear film images into the Guillon categories, in order to detect several categories in each patient. The adequacy of the methodology was demonstrated since it achieves reliable results in comparison with the annotations done by optometrists.