Researchers have developed a conversational diagnosis system that leverages diagnostic knowledge graphs to improve accuracy and efficiency. The system first generates diagnostic hypotheses based on dialogue context and then verifies these hypotheses through clarifying questions. To evaluate the system, they utilized PatientSim, a persona-driven patient simulator, along with patient data from MIMIC-IV, and adapted it to reflect vague symptom reporting from real-world patients. Experiments demonstrated superior diagnostic accuracy and efficiency compared to existing methods, with physician evaluations confirming the simulator's realism and the clinical utility of the generated questions. AI
IMPACT This research could lead to more accurate and efficient AI-powered diagnostic tools, improving patient care and medical training.
RANK_REASON Academic paper detailing a new AI system for conversational diagnosis. [lever_c_demoted from research: ic=1 ai=1.0]
- alphaXiv
- arXiv
- CatalyzeX
- DagsHub
- Gotit.pub
- Hugging Face
- Jeongmoon Won
- MIMIC-IV
- PatientSim
- ScienceCast
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →