An In-Depth Analysis of Methods for Identifying Diabetic Retinopathy

One of the main issues that diabetes patients have with their eyes is diabetic retinopathy. High blood sugar levels in this disorder harm the veins of the retina and may potentially cause temporary or complete vision loss. It is difficult to identify in the initial phases since there are no modest manifestations. However, if left untreated, it can occasionally result in visual impairment and have a significant negative influence on eyesight. Significant visual deterioration is the result of this illness, which most diabetes patients are unaware of. Diabetes-related retinopathy must be diagnosed as soon as possible to prevent further eye complications. Machine learning (ML) has come a long way in recent years and is now an integral part of many medical fields. In this work, a review of diabetic retinopathy identification techniques was done. It can be concluded that CNN models outperformed all the other models in terms of accuracy ranging from 83.05% to 98%.