A Comprehensive Survey on the Detection of Diabetic Retinopathy.

Diabetic Retinopathy (DR) is one of the sight-threatening diseases entered the top 10 diseases that cause death worldwide. There has been tremendous growth in diabetic patients in the last decade due to lifestyle, uncontrolled diet and genetic causes. Nevertheless, the detection of Diabetic Retinopathy in an earlier stage may save the vision of many people. Imaging technologies in the medical field play a highly vital role in understanding, modelling, detecting and diagnosis of diseases. Detection of abnormalities in the retinal image will help to detect, classify and grade the disease. However, it is necessary to segment the retinal image and extract the optic disc, optic cup, retinal blood vessels and macula. The final image interprets the diagnosis of retinopathy diseases, obliging the professional medical practitioner and representatives. This paper describes a comprehensive survey of data acquisition techniques, diabetic retinopathy and the performance of purposed models in this field. The survey involves several papers comprising the research direction of a screening system and lesion segmentation of DR, disease severity grading and abnormality detection. Furthermore, research challenges are discussed and directions are suggested. [ABSTRACT FROM AUTHOR]