A Novel Method of Detecting Neovascularization Regions in Digital Fundus Photographs

Neovascularization is one of the retinal signs we see in the proliferative stage of diabetic retinopathy (DR). Fundus photography is an inexpensive and simple test to identify neovascularization regions when compared to Optical Coherence Tomography and Fluorescein Angiography. Yet, there are no robust methods in the industry that deems to be a benchmark in neovascularization identification. This paper highlights a novel and a robust methodology for neovascularization detection. It incorporates image processing methods and machine learning techniques to segment the regions with neovascularization.