AI-driven early disease detection: Machine learning models for predicting health risks and enhancing healthcare

Abstract

Early disease detection is pivotal for reducing mor- tality rates, yet manual diagnostic processes remain inefficient. This paper proposes a web-based system leveraging ma- chine learning (ML) to predict Alzheimer’s disease, diabetic retinopathy, brain tumors, and heart disease. Key innovations include:1.Multi-disease integration: Combines distinct ML models into a unified platform.2.Explainability via LLMs: Employs large language models (e.g., GPT-4, LLaMA) to generate patient-friendly diagnos- tic rationales.3.Cloud-native architecture: Utilizes React.js (frontend), FastAPI (backend), and AWS (deployment) for scalability.4.Compliance: Adheres to HIPAA and GDPR standards for data security. Experimental results demonstrate accuracies of 85-95% across diseases, with response times under 3 seconds. The system’s modular design allows seamless integration of new diseases, positioning it as a transformative tool for preventive healthcare. Index Terms-machine learning, disease detection, large language models, healthcare, web-based system