Early Detection of Diabetic Retinopathy and Nephropathy using Machine Learning

Tangadapally Sai Kumar1, Kankarla Deepika2, Pureti Sahasra3, Burjukindha Sai vardan goud4,
Mynampati Anjali5, Mr. Ekkaluri kiran kumar6
Student1-5, Dep of CSE, Sphoorthy Engineering College, Hyderabad, India
Assistant Professor6, Dep of CSE, Sphoorthy Engineering College, Hyderabad, India

Abstract :Diabetes mellitus is a chronic disease that leads to severe complications such as Diabetic Nephropathy (DN) and Diabetic Retinopathy (DR), which are major causes of kidney failure and blindness. Early detection of these complications is essential to prevent irreversible damage. Traditional diagnostic methods are time-consuming, costly, and dependent on expert analysis. This paper proposes an intelligent web-based system that integrates Machine Learning (ML) and Deep Learning (DL) techniques for early detection of DN and DR. The system uses clinical parameters for nephropathy prediction and retinal fundus images for retinopathy detection. Among various models tested, XGBoost achieved the best performance for DN prediction, while a Convolutional Neural Network (CNN) was used for DR classification. The models are integrated into a Streamlit-based web application that provides real-time predict ons. The system acts as a decision-support tool, improving early diagnosis, reducing workload, and enhancing healthcare accessibility.
Keywords: Diabetes Mellitus, Diabetic Nephropathy, Diabetic Retinopathy, Machine Learning, Deep Learning, XGBoost,CNN, Healthcare AI