北京启动医疗人工智能应用评测服务

Core Viewpoint - The Beijing Municipal Health Commission has launched an evaluation service for artificial intelligence applications in the medical field, focusing initially on thoracic surgery, marking the first external evaluation service from the national AI application pilot base in healthcare [1] Group 1: Evaluation Service Overview - The evaluation service aims to assess AI applications in healthcare, with a focus on thoracic surgery, and will prioritize promoting outstanding applications to medical institutions [1] - The initiative signifies the expansion and deepening of the national AI application pilot base's external service capabilities in Beijing [1] Group 2: Evaluation Requirements - The evaluation of medical AI must adhere to high standards, emphasizing three key aspects: comprehensive evaluation content, authoritative evaluation data, and scientific evaluation results [2][3] - Evaluation content should consider not only accuracy but also ethics, information security, and the reasoning process behind AI conclusions [2] - The evaluation data must be based on real clinical cases and the latest medical guidelines to accurately reflect AI applications in serious medical scenarios [2] Group 3: Evaluation Methodology - The evaluation center has established a multi-dimensional assessment standard with six core dimensions, including medical compliance, evidence-based knowledge, and diagnostic accuracy, covering over 70 specific evaluation tasks [4] - The evaluation process involves user application, system assessment, and expert review, ensuring a rigorous and objective evaluation [6] - An innovative "referee model" has been introduced to assess the reasoning logic and diagnostic thought processes of AI applications, ensuring a fair and scientific evaluation [6]