Recognition of Diabetic Retinopathy using Deep Learning

Retinopathy is the disorder of the retina. Diabetic Retinopathy (DR) is the primary cause of vision loss and blindness. The diagnosis of DR is a difficult and time-consuming process that necessitates skilled medical professionals. The proposed approach aims to identify a specific stage linked to diabetic retinopathy using deep learning approach. The basic DR stages can be identified by the proposed methodology, which includes Mild NPDR (Non-Proliferative Diabetic Retinopathy), Moderate NPDR, Severe NPDR, PDR (Proliferative Diabetic Retinopathy) and No DR. A dataset from Kaggle is utilized to train the model using a Convolutional Neural Network (CNN). The work is carried out on a Google Colab Notebook using Python programming. The proposed methodology yields training and validation accuracy of 86.66% and 75.96% respectively.