HealthGuard: Early Detection of Chronic Diseases Using Machine Learning

Abstract

The rising incidence of chronic illnesses such as heart disease and diabetes emphasizes the need for accessible, technology-driven tools that support early risk identification and preventive care. This article presents a web-based application that enables users to assess their potential risk for these conditions by entering key clinical parameters, such as blood pressure, glucose level, cholesterol, age, and BMI. The system incorporates two pre-trained machine learning models—an Artificial Neural Network (ANN) for heart disease prediction and an XGBoost classifier for diabetes detection—both capable of analyzing user input to generate real-time, personalized risk evaluations. Developed using  Python and the Flask framework, the platform features a modular architecture that streamlines data validation, preprocessing, prediction, and result visualization. Users interact with a simple, intuitive interface that not only facilitates seamless data entry but also provides actionable insights based on model predictions.  Additionally, the application includes a contact form for feedback or support, enhancing user experience and system responsiveness. Built entirely with open-source technologies, the solution is scalable, lightweight, and easy to maintain, making it ideal for both academic demonstration and deployment in real-world health monitoring scenarios. While it is not a replacement for professional medical consultation, the tool functions as a supportive pre- screening system that raises awareness and encourages timely medical attention, particularly in underserved or remote areas. This project illustrates the impactful use of artificial intelligence in healthcare by offering a cost-effective, user-friendly solution for proactive risk assessment, contributing to improved public health awareness and informed decision-making.