Glaucoma detection through digital processing from fundus images using MATLAB

A considerable proportion of the global population has lost vision due to the late diagnosis of glaucoma, causing ocular pressure on the optic nerve, and blindness to become irreversible. A number of 60.50 million people suffer open angle (OAG) and angle closure glaucoma (ACG) in 2010, increasing to 79.60 million by 2020, and of these, 74% will have OAG. For that reason, the objective of this project was to implement a computational technique of early diagnosis of glaucoma using imaging processing in MATLAB, with the aim of preventing future more severe damages. Together with this, it also sought to improve the precision of the technique to complement the work of the specialist, and providing more accessibility to the patient. The entire ACRIMA database was used for the diagnosis and then analyzed and processed by MATLAB, by extracting the most relevant qualities from the images to reach an optimal diagnosis. The performance test was based on Matlab glaucoma detection proposed method, showing satisfactory results with an accuracy of 94.61%, sensitivity of 94.57% and specificity of 95%. These percentages showed that glaucoma detection using Matlab is feasible.