AI-Driven Chronic Disease Management System

Abstract

The objective of this project is to design and develop an AI-driven Chronic Disease Management System that aids in the early diagnosis and continuous monitoring of chronic illnesses including chronic kidney disease (CKD), Liver Disease, and Diabetes. The system is designed as a cross-platform mobile application built with React Native, featuring secure login/signup screens for both doctors and patients. Machine learning models using the Random Forest Classifier algorithm are integrated to predict disease likelihood based on clinical indicators. These models accept input through manual form-based entry as well as automated extraction from uploaded medical reports using OCR and NLP methods. The extracted data is structured and stored in Firebase and subsequently processed by the disease-specific prediction models to generate risk classifications. For Blood Pressure (BP) monitoring, the system does not use machine learning models. Instead, it implements a statistical anomaly detection approach using standard deviation thresholds. The app evaluates whether new BP values deviate significantly from a patient’s historical average to determine abnormal patterns. This is done using simulated datasets to mimic real-world input for testing and validation. The application offers a dashboard interface for patients to upload reports and visualize their health status, while doctors can view patient data and provide recommendations. The system prioritizes accessibility, privacy, and scalability, supporting chronic disease management through a smart, AI-integrated, mobile-first approach.