This research evaluates the capabilities of large language models in diagnosing rare diseases, addressing a critical gap in clinical support systems.

• Tests LLM performance across a spectrum of rare conditions from Abdominal Actinomycosis to Wilson's Disease
• Examines how well AI can support primary care physicians in making difficult diagnoses from patient conversations
• Addresses the unique challenges of rare disease identification in healthcare settings

This research is particularly valuable for medical professionals seeking to leverage AI as a diagnostic aid for conditions that are inherently difficult to identify due to their rarity, potentially improving patient outcomes through earlier detection and referral.

Rare Disease Differential Diagnosis with Large Language Models at Scale: From Abdominal Actinomycosis to Wilson's Disease