A New Benchmarking for Diabetic Retinopathy Using Machine Learning, Deep Learning and Image Processing Techniques

An elaborated insight into the detection of DR in person eye with the use of different methodologies like image processing, machine learning, and deep learning is discussed. A widely recognized fact is that the level of sugar in blood must be maintained correctly, and its plays an important role in every organ. If it varies abruptly, eventually it will leads to many disease. It is proven that if a person is affected by Diabetics, he has many possibilities of getting affected by Diabetic Retinopathy. As the eyes are in close relation to the brain, any such severe disease affects the retinal connectivity of the eye, it will finally end up with the damage in the brain nervous system, causes the total blindness of the eye, so that the earlier detection of such problems is more needed. The only portion of the human body where blood circulation can be clearly detected is the retina. DR is a disease of the human eye that causeharm to the retina of the eye which can eventually result in total blindness. There are several methods for segmenting the blood vessels in the retina, just the once the retinal nerve fibres are segmented, it is possible to determine whether the eyes are affected by diabetic retinopathy or not. Even the Diabetic Retinopathy detection depends on the Retinal Nerve Fibre Layer network's coverage area. If the entire area of the nerve fibre is less than total area of the nerve network, the eyes are strickenby DR, anywhere the total area of the nerve network is greater, the eyes are not affected by diabetic retinopathy and therefore are usual. The retinal blood vessels can be impaired by many systemic diseases, which make retinal image analysis a possible diagnostic method, since it enables simple and non-invasive assessment of vascular changes. To prevent total loss of sight, timelyrevealing of DR is needed. For the diagnosis of diabetic retinopathy, physical measures such as visual acuity tests, pupil dilation and optical accuracy tomography are used. In terms of time, however it is expensive and could impact patients. This article provides a brief overview of how machine learning, deep learning, and image processing techniques can be used to detect DR in human eyes. As well as they will gain the knowledge about the disease in the human eye. To put it more simply, the paper is written as an introduction to diabetic retinopathy and the different methodologies tried to cure this disease. Until now, various research analysts have whittling away at this title's theme. To begin that, a large number of research papers were gathered from different sources, reviewed in depth, and a brief overview of eye disease issues was produced and presented here in a precise manner. In this context, latest works done by different writers, all across the world is being discussed in this context so that any researcher working in the field of ophthalmology can identify their own new research challenge using this review article as a guide.