Diabetic Retinopathy is a severe diabetes complication that can cause vision loss or blindness. Image analysis has been increasingly used by medical researchers to accurately diagnose diseases. As a result, a computational model was developed to predict the presence of Diabetic Retinopathy (DR) using retinal images and a neural network. The computational model consists of two stages: feature extraction and classification. In the feature extraction phase, digital fundus images were analyzed to identify the most relevant features such as Blood Vessels and Micro aneurysms. The study was conducted using data from the Diabetic Retinopathy dataset available on Kaggle Community. Our objective is to use deep neural networks to automatically differentiate between healthy and pathological retinal fundus images. This is because deep learning is a highly effective machine learning method that has been proven to be extremely accurate in various computer vision tasks. To achieve this goal, we have employed convolutional neural networks (CNN) in our study to classify retinal images as either healthy or unhealthy. Finally, a convolutional neural network (CNN) was used to predict Diabetic Retinopathy. The proposed methodology achieved a 95.41% accuracy rate based on the model's results.

Diabetic Retinopathy, Retinal Images, Digital Fundus Images, Blood Vessels, Micro Aneurysms, CNN and Deep Learning