Assessing Changes in Diabetic Retinopathy Caused by Diabetes Mellitus and Glaucoma Using Support Vector Machines in Combination with Differential Evolution Algorithm

The aim of this study is to evaluate the changes related diabetic retinopathy (DR) (no changes, small or moderate changes) in patients with glaucoma and diabetes using artificial intelligence instruments: Support Vector Machines (SVM) combination a powerful optimization algorithm—Differential Evolution (DE). In order classify DR make predictions various situations, an approach including SVM optimized DE was applied. role optimizer automatically determine parameters that lead lowest classification error. conducted on sample 52 patients: particularly, 101 eyes mellitus, Ophthalmology Clinic I “St. Spiridon” Clinical Hospital Ia?i. criteria considered modelling action were normal hypertensive open-angle glaucoma, intraocular hypertension associated diabetes. other types pseudoexfoliation, pigment, cortisone, neovascular primitive angle-closure, those without diabetes, excluded. assessment carried out Volk lens Fundus Camera Zeiss retinal photography dilated pupil, inspecting all quadrants. for classifying (early treatment study—ETDRS) were: (absence DR), mild forma nonproliferative (the presence single micro aneurysm), form (micro aneurysms, hemorrhages 2–3 quadrants, venous dilatations soft exudates quadrant), severe dilatation quadrants) proliferative (disk neovascularization different quadrants). Any new clinical element occurred subsequent checks, which led their inclusion forms retinopathy, be result progression retinopathy. results obtained very good; testing phase, 95.23% accuracy has been obtained, only one being wrongly classified. effectiveness algorithm (SVM), developed optimal DE, used demonstrated.