ANALYSIS OF HUMAN RETINAL MORPHOLOGY FOR THE IDENTIFICATION OF DIABETIC RETINOPATHY AND MACULOPATHY USING FUNDUS IMAGES

Abstract

Diabetic Retinopathy (DR) is a microvascular inconvenience of newlinediabetes, bringing on anomalies in the retina and in the most pessimistic newlinescenario, blindness or lack of sight. The word retinopathy shields several newlinedisorders of the retina, which can distress vision. Retinopathy typically owes newlineharm to the tiny blood vessels in the retina. It can be either caused by newlinediabetes or other ailments such as very high blood pressure. newlineD iabetic Retinopathy (DR) could be a frequent trouble of diabetes. newlineIndeed, it is the leading reason behind visual defect or blindness among the newlineworking individuals. Despite the health care state of affairs, individuals with newlinediabetes are twenty five times more likely to develop visual defect compared newlinewith people who do not suffer from diabetes. Diabetic Retinopathy could be a newlinesilent disease; as a result, it should solely be recognized by the patient newlinebefore the changes within the retina have progressed to the level wherever newlinetreatment is impossible and much complicated. newlineIn numerous conditions, the digital image processing methods can newlinebe used to identify or track the changes in the retinal region. The newlineclassification of abnormalities within the human eye using these techniques newlinehas increased progressively in the recent past. The early diagnosis and newlinetreatment may prevent or decrease the vision loss of the patient. Hence, the newlineaim of this study is to develop computer aided diagnostic system (CAD) newlinecapable of supporting medical decision, improving diagnosis quality and newlinedecreasing the workload of medical professionals within the background of newlinediagnose diabetic retinopathy. The novel framework consists of the following newlinestages: (1) Extract blood vessels, (2) Extract optic disk, (3) Detection of newlineexudates from retinal digital fundus images and (4) Classification. newlineThe samples or dataset for this study has been obtained from a newlineleading eye care hospital, Chennai, India. The computer aided diagnostic newlinesystem (CAD) based on morphological features in combination with SVM newlineclassifier is proposed which enables diagnosis o