A Systematic Literature Review on Diabetic Retinopathy Stage Diagnosis Using Deep Learning Approach

Abstract

Diabetic retinopathy is a medical condition that occurs due to diabetic mellitus causing blindness in the later stages. As the name says it is caused for people with the diabetes of type 1 and type 2 for a period of 20 years or more. However, early diagnosis and proper treatment at the initial stages could reduce aggressive forms of threatening retinopathy by up to 90%. According to the 2019 census from the ninth edition of the IDF Diabetes Atlas, India is put in the second slot right up to 2045 with people who are aged between 20-79 years. In 2019, India had 77 diabetic patients. In the recent survey of 2022 by the Union Health Ministry of India, almost 17% of the Indian population has diabetic retinopathy of which 3.6% is sight threatening. Specific to the rural population, 10.4% had diabetes of which 10.3% were diabetic retinopathy affected. From the WHO survey, it is estimated that by 2030, diabetics with retinopathy will increase by 50%. Without proper treatment, the individual is more likely to lose his vision. So, to prevent it, early detection of diabetic retinopathy and its stage is necessary. Key Words: Diabetic retinopathy, fundus, proliferative and non-proliferative, deep learning.