首个医学多模态可解释性基准，评测AI能否用患者能懂的话描述影像结果。
MEDLAYXPLAIN: Benchmarking the Expert-Lay Gap in Medical Vision-Language Models
- 构建分层语义框架，用医疗本体映射实现患者语言精准重构
- 12万+图像样本验证：现有模型在通俗表达上严重退化
- 轻量级评估器突破传统指标局限，贴近临床判断
医学视觉语言模型（Med-VLMs）在专家级表现优异，但其生成患者可理解描述的能力尚未充分探索。随着《21世纪治愈法案》要求立即向患者开放影像诊断结果，评估这些模型是否能弥合专家与患者之间的语言鸿沟，已成为关乎患者教育和共同决策的紧迫问题。为此，我们提出MedLayXPlain，首个大规模多模态基准与评估框架，用于医学通俗语言生成（MLLG）。MedLayXPlain-122K包含来自12个公开数据集的122,789个区域锚定样本，覆盖8种成像模态，每条样本配有专家与通俗描述，均基于三层次统一医学语言系统（UMLS）本体层级，涵盖7个语义组、43个语义类型和2,411个医学概念。通俗描述通过分层本体验证精炼（HOVER）流程生成，包括患者中心词汇映射、大模型约束重写及跨模型视觉验证，确保语义等价且避免幻觉。我们进一步提出MedLayEval，一个从270亿参数验证器蒸馏出的30亿参数轻量级评估器，从五个临床相关维度评分专家-通俗一致性，解决了标准自然语言生成指标与临床判断的相关性差的问题。在MedLayXPlain-122K上对33个视觉语言模型的评测显示，存在系统性专家-通俗差距：医学专用模型在专家描述上表现良好，但在通俗表达上显著退化；通用模型虽更易懂，却缺乏临床精确性，证实当前范式均无法有效支持面向患者的沟通。
原文摘要 · Abstract (English)
Medical Vision-Language Models (Med-VLMs) achieve strong expert-level performance, yet their ability to generate patient-accessible descriptions remains underexplored. With the 21st Century Cures Act now mandating immediate patient access to diagnostic imaging results, evaluating whether Med-VLMs can bridge this Expert-Lay Gap is both urgent and clinically consequential for patient education and shared decision-making. To this end, we introduce MedLayXPlain, the first large-scale multimodal benchmark and evaluation framework for Medical Lay Language Generation (MLLG). MedLayXPlain-122K provides 122,789 region-grounded samples across 8 imaging modalities from 12 publicly available source datasets, each comprising a medical image with paired expert and lay captions anchored in a three-level Unified Medical Language System (UMLS) ontology hierarchy spanning 7 semantic groups, 43 semantic types, and 2,411 medical concepts. Lay captions are constructed via Hierarchical Ontology-Verified Refinement (HOVER), a three-step pipeline combining patient-centric vocabulary mapping, LLM-based constrained rewriting, and cross-model visual verification to enforce semantic equivalence while preventing hallucination. We further introduce MedLayEval, a lightweight 3B evaluator distilled from a 27B verifier that scores expert-lay alignment across five clinically grounded attributes, addressing the poor correlation between standard NLG metrics and clinical judgment. Benchmarking 33 VLMs on MedLayXPlain-122K reveals a systematic Expert-Lay Gap: medical VLMs achieve strong expert captioning but suffer significant lay-register degradation, while general-purpose VLMs produce more accessible language yet lack clinical precision, confirming that neither current paradigm adequately serves patient-facing communication.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。