Reproduce “MedSimAI: Simulation and Formative Feedback Generation to Enhance Deliberate Practice in Medical Education” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Yann Hicke, Jadon Geathers, Kellen Vu, Justin Sewell, Claire Cardie, Jaideep Talwalkar, Dennis Shung, Anyanate Gwendolyne Jack and 3 more
MedSimAI, a multi-institutional collaboration led by Cornell University, provides an AI-powered simulation platform for medical education that generates realistic patient interactions and automated, structured feedback. The system demonstrated a 6.0-point improvement in OSCE history-taking scores in one integrated curriculum and achieved 87.0% thresholded accuracy in identifying proficiency levels compared to human evaluators.
Key insights
- Effective integration of AI-powered simulation tools into the curriculum, rather than mere availability, is crucial for measurable performance improvements.
- LLM-based automated scoring can reliably identify student proficiency thresholds (87% accuracy), providing a scalable tool for formative assessment and triage.
- A multi-phase co-design approach with medical education experts ensures that AI platforms are pedagogically aligned and address real-world training needs.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.