Diabetic Retinopathy Identification Depend On Deep Learning Techniques: A comparative Study

Diabetic Retinopathy (DR) is a prevalent complication of diabetes that affect the retina. If not detected early, it can cause loss of vision. Diabetic Retinopathy is considered to be the cause for vision loss to patients with diabetes, mostly the working-age people in developing nations. Unfortunately, it is an irreversible and treatment can only help sustain vision. When DR detected and managed in its early stages this can minimize the risk of blindness. DR manual diagnosis using retina images by doctors is time-consuming, labor-intensive, and costly. Deep learning has emerged as a widely used technique, has proven good result in analyzing retinal fundus images, offering improved performance. Convolutional neural networks are highly effective in medical image analysis as a deep learning method. This article reviews and analyses some methods for detecting and classifying DR color fundus images using deep learning models. Additionally, it examines some of the datasets for DR retina images.