Doctorina MedBench：基于代理的医疗人工智能的端到端评估

Doctorina MedBench presents a comprehensive evaluation framework designed for agent-based medical AI, focusing on simulating realistic physician-patient interactions. Unlike traditional medical benchmarks that rely on standardized test questions, this approach models a multi-step clinical dialogue where either a physician or an AI system must gather medical history, analyze supplementary materials (including lab reports, images, and medical documents), formulate differential diagnoses, and provide personalized recommendations. System performance is assessed using the D.O.T.S. metric, which consists of four components: Diagnosis, Observations/Investigations, Treatment, and Step Count, enabling a thorough evaluation of clinical communication effectiveness.