Enhancing Early Diagnosis and Precise Classification of Type 2 Diabetes with Diabetic Retinopathy Using Machine Learning and Image Processing

The paper describes an accurate classification of Type 2 Diabetes with Diabetic Retinopathy using Machine Learning and Image Processing. The objective of the paper is to find Diabetic Retinopathy from the image of retina whether a patient has the disease if yes, is it on the first stage, moderate or severe stage. The Dataset used in the study are Diabetic Retinopathy, Aptot2019 _blindness detection and Messidor2. The research shows the comparison between CNN models, including Xception, Alex Net, VggNet-s, VggNet-16, and ResN et-50. The ResN et-50 shows better performance than other models. The research ensured 96 % accuracy, and the value of precision is 0.97 from the proposed but it can be increased by improving the model. The study shows an excellent result with deep learning but there can be more improvement in the quality of images and fine-tuning by using other optimal technics.