DrVD-Bench：视觉语言模型在医学图像诊断中是否像人类医生一样推理？
DrVD-Bench: Do Vision-Language Models Reason Like Human Doctors in Medical Image Diagnosis?
AI总结：
提出首个临床视觉推理多模态基准DrVD-Bench，包含7,789个图像-问题对，覆盖20种任务和5种成像模态，系统评估19个VLMs发现其随推理复杂度增加性能骤降，仍依赖捷径而非扎实视觉理解。
摘要：
视觉语言模型（VLMs）在自然图像上展现出强大的零样本泛化能力，并在可解释的医学图像分析中显示出初步前景。然而，现有的基准并未系统地评估这些模型是否真正像临床医生那样推理，还是仅仅模仿表面模式。为了解决这一差距，我们提出了DrVD-Bench，这是首个用于临床视觉推理的多模态基准。DrVD-Bench包含三个模块：视觉证据理解、推理轨迹评估和报告生成评估，共计7,789个图像-问题对。我们的基准涵盖20种任务类型、17个诊断类别以及五种成像模态——CT、MRI、超声、放射摄影和病理学。DrVD-Bench的结构明确反映了从模态识别到病灶识别和诊断的临床推理工作流程。我们对19个VLMs进行了基准测试，包括通用和专用医疗、开源和专有模型，并观察到随着推理复杂性的增加，性能急剧下降。虽然一些模型开始展现出类人推理的痕迹，但它们通常仍然依赖于捷径相关性，而非基于扎实的视觉理解。DrVD-Bench提供了一个严谨且结构化的评估框架，以指导临床可信VLMs的开发。
英文摘要：
Vision-language models (VLMs) exhibit strong zero-shot generalization on natural images and show early promise in interpretable medical image analysis. However, existing benchmarks do not systematically evaluate whether these models truly reason like human clinicians or merely imitate superficial patterns. To address this gap, we propose DrVD-Bench, the first multimodal benchmark for clinical visual reasoning. DrVD-Bench consists of three modules: Visual Evidence Comprehension, Reasoning Trajectory Assessment, and Report Generation Evaluation, comprising a total of 7,789 image-question pairs. Our benchmark covers 20 task types, 17 diagnostic categories, and five imaging modalities-CT, MRI, ultrasound, radiography, and pathology. DrVD-Bench is explicitly structured to reflect the clinical reasoning workflow from modality recognition to lesion identification and diagnosis. We benchmark 19 VLMs, including general-purpose and medical-specific, open-source and proprietary models, and observe that performance drops sharply as reasoning complexity increases. While some models begin to exhibit traces of human-like reasoning, they often still rely on shortcut correlations rather than grounded visual understanding. DrVD-Bench offers a rigorous and structured evaluation framework to guide the development of clinically trustworthy VLMs.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
