Abstract: Diabetic retinopathy (DR) is one of the most serious complications of diabetes, which affects the retina and can lead to impaired vision or blindness if detected and treated late. The project develops an A. I. based diagnostic system whose SDL with classified grading for DR identification using retinal fundus images. The system uses CNNs-based medical image analysis and divides it into five different classes: NoDR, Mild, Moderate, Severe, and ProliferativeDR. Image preprocessing enhances the extraction of features and prediction accuracy. To further improve accessibility, the project includes a web application based on Flask which allows users to upload their images and instantly receive a diagnostic report. The development also reduces medical professionals' workload while providing a scalable way to massively address the rising tide of diabetes-related blindness, thus making quality eye care available everywhere.
Raj et al. (2024) studied this question.