Patient Simulation

Patient simulation research focuses on creating realistic virtual patients for medical training, research, and model evaluation. Current efforts leverage large language models (LLMs) coupled with knowledge graphs and diffusion transformers to generate diverse patient presentations, including symptoms, medical histories, and even medical images, within interactive conversational frameworks. This allows for the assessment of LLMs in complex clinical scenarios and the development of more effective training tools for healthcare professionals, ultimately improving medical education and potentially optimizing clinical decision-making.