Performance analysis of automated screening system for diabetic retinopathy detection using retinal fundus image analysis

Abstract

This thesis intends to compose an instinctive system which is capable of sensing the abnormalities in anatomical features obtained from retinal images Many eye diseases are prominent around the world nations and diabetes is the major cause of such complications Diabetes mellitus leads to vision affecting condition which is medically known to be diabetic retinopathy This disease has become responsible for 12 of blindness case every year in United States alone The calming truth is nearly 90 of new cases can be averted with regular checkups proper treatment and monitoring the eye With limited resources equipments and rarity of specialist demand the innovative approach to standardize the procedure with assistance of specialized software in a computerized environment Ophthalmologists require the colour fundus images of patients for monitoring the retinal images of patients It is said that the number of patients is extremely high in the working age group Now it becomes obvious that having this number at an alarming rate the regular screening images will demand immediate and more attention from ophthalmologists to prevent vision lost. newline