Predicting Diabetic Retinopathy and Nephropathy Complications Using Machine Learning Techniques

Diabetic Retinopathy (DR) and Diabetic Nephropathy (DN) are serious complications associated with diabetes, leading to severe health outcomes such as blindness and kidney failure. Early prediction and intervention are crucial in preventing or delaying the progression of these complications. This project aims to predict the likelihood of DR and DN using machine learning techniques. The dataset used for this study includes key health parameters such as age, diabetes duration, HbA1c levels, insulin usage, smoking habits, and more. The machine learning models used for prediction include Bagging Classifier, Adaboost, and Gradient Boosting, which are trained to identify the risk of these complications based on the input data. The project is developed as a web-based application using the Flask framework, allowing users to enter their health data and receive predictions about their risk for DR and DN. The system includes essential features like user registration, login, and a classification module that performs the prediction. This project not only aims to provide an accessible tool for healthcare practitioners but also empowers individuals to monitor their health conditions. By automating the prediction process, this system can help identify high-risk individuals early on, allowing for timely intervention and better management of diabetes. The models’ performance will be evaluated based on accuracy, precision, recall, and F1-score to ensure reliable predictions. This approach integrates machine learning with a web interface to offer an efficient and user-friendly solution for predicting diabetic complications.

Keywords: Diabetic Retinopathy, Diabetic Nephropathy, Machine Learning, Bagging Classifier, Adaboost, Gradient Boosting, Diabetes, Prediction, Healthcare