当模型学会问为什么：面向可信医疗视觉语言模型的自适应因果推理
When Models Learn to Ask Why: Adaptive Causal Reasoning for Trustworthy Medical Vision-Language Models
AI总结：
提出MedCausalX框架，通过因果推理链、自适应反射架构和轨迹级因果校正，解决医疗VLM中的虚假相关和推理不一致问题，显著提升诊断一致性和减少幻觉。
摘要：
视觉语言模型（VLM）通过整合视觉感知与语言推理，实现了可解释的医疗诊断。然而，现有的医疗思维链（CoT）模型缺乏明确表示和执行因果推理的机制，使其易受虚假相关影响，限制了临床可靠性。我们指出了医疗CoT推理中的三个核心挑战：如何自适应触发因果校正、构建高质量因果-虚假对比样本、以及保持推理轨迹的因果一致性。为应对这些挑战，我们提出MedCausalX，一个显式建模医疗VLM中因果推理链的端到端框架。我们首先引入CRMed数据集，提供细粒度解剖标注、结构化因果推理链和反事实变体，引导学习超越表面相关性的因果关系。基于CRMed，MedCausalX采用两阶段自适应反射架构，配备⟨causal⟩和⟨verify⟩标记，使模型能够自主决定何时以及如何进行因果分析和验证。最后，通过错误归因强化学习优化的轨迹级因果校正目标，细化推理链，使模型能够区分真正的因果依赖与捷径关联。在多个基准上的大量实验表明，MedCausalX持续优于最先进方法，诊断一致性提升+5.4分，幻觉减少超过10分，并达到最高的空间定位IoU，从而为因果基础的医疗推理设立了新标准。代码和数据集可在https://github.com/zhcz328/MedCausalX获取。
英文摘要：
Vision-Language Models (VLMs) have enabled interpretable medical diagnosis by integrating visual perception with linguistic reasoning. Yet, existing medical chain-of-thought (CoT) models lack explicit mechanisms to represent and enforce causal reasoning, leaving them vulnerable to spurious correlations and limiting their clinical reliability. We pinpoint three core challenges in medical CoT reasoning: how to adaptively trigger causal correction, construct high-quality causal-spurious contrastive samples, and maintain causal consistency across reasoning trajectories. To address these challenges, we propose MedCausalX, an end-to-end framework explicitly models causal reasoning chains in medical VLMs. We first introduce the CRMed dataset providing fine-grained anatomical annotations, structured causal reasoning chains, and counterfactual variants that guide the learning of causal relationships beyond superficial correlations. Building upon CRMed, MedCausalX employs a two-stage adaptive reflection architecture equipped with $\langle$causal$\rangle$ and $\langle$verify$\rangle$ tokens, enabling the model to autonomously determine when and how to perform causal analysis and verification. Finally, a trajectory-level causal correction objective optimized through error-attributed reinforcement learning refines the reasoning chain, allowing the model to distinguish genuine causal dependencies from shortcut associations. Extensive experiments on multiple benchmarks show that MedCausalX consistently outperforms state-of-the-art methods, improving diagnostic consistency by +5.4 points, reducing hallucination by over 10 points, and attaining top spatial grounding IoU, thereby setting a new standard for causally grounded medical reasoning. The code and dataset are available at https://github.com/zhcz328/MedCausalX.
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。
