Design of Artificial Intelligence based Teaching Assistance System for Online "Basic Disaster Rescue Medicine" Courses

Disaster rescue medicine is an integral component of emergency medical education, with the goal of educating students in the necessary knowledge and life-saving skills used during emergency situations. The current educational methods are plagued by several drawbacks such as inflexible content delivery, absence of real-time interaction, limited remote access, and insufficient testing of practical decision-making abilities. The conventional systems tend to be static and do not involve learners in interactive, real-time rescue contexts. In order to tackle these challenges, this work suggests a Web-Enabled BERT-based Teaching Assistance System specifically designed for the "Basic Disaster Rescue Medicine" course. The system architecture consists of modular content delivery with web technologies, simulation-based interactive case practice, real-time skill testing, and an intelligent feedback loop that adjusts learning materials according to user performance. The major innovations encompass dynamic rendering of learning material, asynchronous support for remote learners, and inbuilt behaviour tracking to monitor performance and decision validity. Experimental evaluation of the platform included comparative testing against standard learning environments. Experimental findings establish a significant improvement in learner interest, knowledge retention, and emergency response preparedness. Performance measures comprise accuracy at 97.8%, precision at 97.8%, recall at 97.5%, and F1-score at 97.5%, which signify high efficacy of the proposed system. This web-based system is a breakthrough towards scalable, interactive, and result-oriented training in emergency medical education with high potential to be further enriched through AI-based adaptation and immersion simulations.