Application of an iterative generative AI-augmented teaching model in medical immunology education

Introduction: Immunology education faces persistent challenges, including abstract concepts, theory-practice disconnect, and delayed feedback. Generative AI offers opportunities, yet its integration into pedagogical frameworks remains underexplored. Methods: We developed an Iterative AI-Augmented Immunology Education (IAIE) model integrating generative AI with the 4C/ID model and cognitive load theory. A 12-week quasi-experimental study enrolled 177 medical undergraduates (experimental: n = 88; control: n = 89), with outcomes assessed using the Host-Pathogen Interactions Concept Inventory (HPI-CI) and a modified Clinical Decision-Making in Nursing Scale (CDMNS). Results: The experimental group showed significantly greater knowledge mastery (82.7 vs. 70.1, Cohen's d =1.12) and clinical decision-making accuracy (+31.7%). Over 84% of students recognized AI's value in understanding complex mechanisms. Reflective practice scores indicated high levels of reflective ability, and DREEM assessments confirmed strong student acceptance of the learning environment. Discussion: The IAIE model effectively bridges theory and practice through AI-driven scaffolding that modulates cognitive load, fostering higher-order thinking and critical engagement with AI-generated content. This theory-informed framework offers a transferable approach for competency-based medical education.