Figure 4 and Figure 6 Raw Data for Paper: "Simulated patient systems powered by large language model-based AI agents offer potential for transforming medical education"

Yu et al. introduce AIPatient, a virtual patient that uses modern AI and a curated medical knowledge base to hold realistic, personality-aware clinical conversations. The system shows strong accuracy in question answering and medical-term recognition, with readable, reliable, and consistent outputs for training and decision support. The datasets in this repository contains (i) raw bootstrap outputs used to generate Figure 4 (per-replicate metrics [F1, Precision and Recall] across entity categories and models for 10,000 resamples), and (ii) medical-student scoring data for Figure 6, standardized to 20 students (D1–D20) with four evaluations each (2 AI, 2 Human). These files support reproducibility of the AIPatient project’s performance figures and user-study analyses.