Design and Development of Diabetic Retinopathy by Using Convolutional Neural Network Classifier

These days, diabetes is a fairly common condition in humans. A group of diseases known as diabetes mellitus are related to elevated blood sugar levels in the body. Diabetes may make someone more susceptible to acquiring DR. A disorder affecting the retina's blood vessels is called diabetic retinopathy. Three quarters to forty percent of the estimated 347 million people have diabetic retinopathy. Since this is an invisible issue, annual check-ups are necessary to help control the disease in its early stages and ensure that therapy is effective. This study suggests a treatment for diabetic retinopathy, a condition that affects working-age adults and progresses quickly, sometimes resulting in blindness. Deep learning research is being used to categorize diabetic retinopathy will also help solve real-world problems. Using deep learning approaches that can identify patterns and categorize retinal fundus images as normal or sick, the current study tackles this problem. 3662 retinal pictures from various diabetic retinopathy groups were examined in this study. The literature now classifies diabetic retinopathy into four categories: mild, moderate, severe, and proliferative. Implementing and comprehending how proliferative and non-proliferative diabetic retinopathy differs in the pre-trained model's performance is aim of this study.