AI-ENABLED HEALTHCARE INFRASTRUCTURE FOR REAL-TIME MONITORING AND CLINICAL INTELLIGENCE
#1 Mr. E. Subramanian, @2Siddarth D, @3Sudarsan C, @4 Surya R
#1Assistant Professor, Sri Shakthi Institute of Engineering and Technology, Coimbatore,
esubramaniancse@siet.ac.in,
@2, @3, @4 UG Student, Sri Shakthi Institute of Engineering and Technology, Coimbatore, siddarthd22cse@srishakthi.ac.in sudarsanc22cse@srishakthi.ac.in
suryar22cse@srishakthi.ac.in
Abstract - A comprehensive full-stack healthcare management system built on FastAPI (Python) and Next.js 14 (TypeScript), backed by MongoDB, supporting four stakeholder roles patients, doctors, pharmacists, and administrators through 50+ RESTful API endpoints and a real-time WebSocket interface. Security is enforced via JWT-based RBAC, with HIPAA-aligned compliance maintained through a custom SHA-256 blockchain audit trail. The AI subsystem integrates LLMs (Google Gemini with fallback to Perplexity/OpenAI) for a multilingual health assistant, LLM-driven prescription validation (drug interactions, contraindications, dosage anomalies), and predictive risk analysis for diabetes, hypertension, and cardiovascular disease using scikit-learn Random Forest classifiers trained on an 18-factor patient profile. The pharmacy module manages the full prescription lifecycle including pharmacist–doctor escalation, dispensing, and GST- inclusive invoicing, complemented by appointment scheduling with automatic waitlisting. The system is containerized via Docker Compose (MongoDB, FastAPI, Next.js, Redis), delivering a scalable clinical solution that reduces administrative overhead while improving diagnostic accuracy and medication safety.
Keyword- FastAPI; Next.js; JWT Authentication; Large Language Model; Prescription Validation; Predictive Risk Analysis; Random Forest Classifier; Clinical Decision Support; Role-Based Access Control; Real-Time Monitoring; Electronic Health Records; Drug Interaction Detection; Multilingual Health Assistant; Docker; Artificial Intelligence; Healthcare Information System
