AI-Enhanced Virtual Nursing Systems: Revolutionizing Patient Education in Modern Healthcare

Traditional patient education relies on static brochures, lacking interactivity and increasing healthcare providers' workload. This study introduces a virtual nurse dialogue system using ASR and a fine-tuned LLama3 8B LLM trained on Traditional Chinese medical texts. A RAG framework with a structured knowledge base enhances accuracy and minimizes misinformation. The system features an RWD-based interface with 3D animations and synchronized voice outputs for an immersive experience. Experimental results show 95% accuracy and responses under 5 seconds, with 90% user satisfaction, demonstrating its potential to enhance patient education and reduce healthcare burdens.