Screening of Ocular Diseases Through Retinal Vascular Diameter Measurements

The automated diagnosis and screening of ophthalmologic diseases rely on the examination of retinal vessels morphological features and the exploration of their branching patterns. Our study introduces a new method for automatically identifying the retinal vessel diameter, centerline, as well as the location of Bifurcation and Intersection Points. In this research, we present and validate our method, which provides an automated and innovative approach. Initially, we utilize the skeletonization method to extricate the central line of retinal vasculature. This extracted line is then used to determine the locations of specific points within the vessels. Subsequently, we identify the locations of bifurcations and intersections. In order to achieve precise vessel diameter measurements, we introduce a dependable technique that incorporates a suggested mask to establish the early vascular skeleton alignment. Finally, the vessel's diameter is computed by utilizing the obtained centerline and directional information of the vessel. The proposed method exhibited superior performance in comparison to previous approaches, yielding precise and promising results. There is potential for further development of this technology into an automated system capable of assessing arterio-venous ratio and quantifying arterio-venous nicking in retinal photographs which aids in the identification of individuals at a heightened risk of cardiovascular events, who would benefit from early intervention.