AI in Healthcare Education: What We Learned at AMEE 2024 Conference in Basel

September 13, 2024

A few weeks ago, AIBODY had an invaluable opportunity to present our AI-powered solutions for medical education at the AMEE 2024 Conference and AI Education Symposium in Basel, organized by the International Association for Health Professions Education (AMEE). We took part in a broad range of presentations, networking sessions and workshops focused on AI’s ability to transform medical education and enhance student learning, and explored the latest data on emerging applications of AI in healthcare education and research.

Early Evidence for AI in Medical Education

Institutions are increasingly looking to integrating AI into their curricula to address challenges like limited active learning hours, faculty shortages, and the complexities of healthcare programs. In healthcare education, where large cohorts often experience inconsistencies in learning, AI offers the potential to standardize and improve student outcomes. By boosting teaching capacity and reducing educator workloads, AI can fostergreater student engagement and confidence.

However, its success will rely on aligning AI use with sound educational principles and mitigating risks, especially for clinical learners, who stand to benefit most but are also vulnerable to inaccuracies.

AI's strengths—such as aiding in skills assessment, explaining complex concepts, and managing large amounts of data—make it a valuable tool for both learners and educators. While companies at AMEE presenting AI-driven products have focused on areas like communication skills and clinical reasoning, a critical gap remains: most platforms lack integration with foundational sciences like physiology, a vital element in healthcare education. Filling this gap is essential to creating a more comprehensive, effective learning environment.

Early evidence presented at AMEE shows that AI integration can increase teaching capacity, reduce educator workload, and boost student engagement and confidence. This is particularly relevant for clinical learners, who gain the most from AI support but are also the most vulnerable to AI inaccuracies. Mitigating these risks will be key to ensuring AI's success in education.

Key Themes and Takeaways

• AI has the potential to revolutionize medical education, but should support rather than replace human educators and clinical reasoning	• Collaboration between AI developers, medical educators, and clinical practitioners is critical for successful and ethical AI implementation
• AI applications are being explored for personalized learning, assessment and feedback, case generation, and clinical simulations	• More fundamental research focusing on empiricism rather than technology is needed on the effective integration of AI into pedagogy and practice to ensure it enhances learning outcomes

AI Symposium Highlights

Renowned speakers like Martin G. Tolsgaard, Martin Pusic, Steph Smith, and Ruben Hassiddiscussed AI's transformative power and challenges in health professions education.
World Café sessions focused on AI for learning enhancement, ethical considerations, biasmitigation, and the impact on scholarship and research.
The symposium introduced the International Advisory Committee on AI (IACAI) to guide AI development and implementation in medical education.

Clinical Reasoning Symposium

AI has limitations in clinical practice, such as low specificity and the need for large samplesizes in randomized control trials.
AI is susceptible to bias, similar to human diagnosticians, particularly when presented withsalient distracting features.
AI should support clinical reasoning by making critical diagnostic knowledge available tophysicians.
First impressions play a crucial role in clinical reasoning and influence subsequent diagnosticdecisions.
Large databases of student interactions are needed to improve AI's accuracy in clinicalreasoning assessment.

Networking Session

Highly interactive discussions facilitated by Azer Anatka, Marcus Roquez, and RungrojAngwatcharaprakarn.
Topics included using AI in educational practice, AI in assessment, ethical issues with AI, andAI's impact on the educator role.
AI is seen as a tool to enhance student learning, but should not be the sole focus of education.
Transparency about AI use and limits, as well as responsible and ethical implementation, arekey considerations.

Short Communications

Promising AI applications were showcased for skill assessment (e.g., CPR), communicationtraining, qualitative research analysis, student evaluation, and prompt engineering training.
Benefits include scalability, focused practice and feedback to complement humaninstructors.
Challenges include ensuring transparency about AI use and limitations, as well as effectiveintegration with pedagogy and context.

Warm regards,

Dr Dr Aaron Smith and Edmund White