PREDICTION OF CHRONIC KIDNEY DISEASE DETECTION USING MACHINE LEARNING

Abstract

Chronic Kidney Disease (CKD) is a progressive and life-threatening medical condition characterized by the gradual loss of kidney function over time. Early detection and accurate staging are essential for preventing severe complications such as kidney failure, cardiovascular disease, and mortality. Traditional diagnostic approaches rely on laboratory tests and estimated Glomerular Filtration Rate (eGFR) calculations such as MDRD and CKD-EPI formulas; however, these approaches often fail to incorporate diverse clinical parameters and patient history for effective prediction. Recent advancements in machine learning have enabled the development of intelligent healthcare systems capable of analyzing large clinical datasets to support early diagnosis and risk prediction. This study proposes a machine learning-based system for CKD stage detection and risk prediction using patient clinical attributes such as age, blood pressure, serum creatinine, hemoglobin, blood urea, and comorbidity indicators like diabetes and hypertension. The system integrates a web-based architecture consisting of a Spring Boot backend for authentication and data management, a FastAPI-based machine learning service for model training and prediction, and a React-based frontend for interactive user access. An Artificial Neural Network model is employed to classify CKD stages from Stage 1 to Stage 5 and generate risk scores along with explainable outputs. Feature preprocessing techniques such as data scaling, encoding, and normalization improve prediction accuracy. Additionally, explainable AI techniques help clinicians understand the factors influencing predictions. The proposed platform allows patients to upload laboratory reports, doctors to monitor patient health trends, and administrators to retrain models using new datasets. The system demonstrates how machine learning and modern web technologies can be combined to create an intelligent, scalable, and secure decision support system for early CKD detection and clinical assistance.