arXiv 2605.266212026-05-26cs.CVcs.AI

MedVol-R1：基于奖励驱动的证据基础用于体积推理分割

MedVol-R1: Reward-Driven Evidence Grounding for Volumetric Reasoning Segmentation

School of Biological Science and Medical Engineering, Beihang University, Beijing, China（生物科学与医学工程学院，北京航空航天大学）；Center for Information and Computer Science, School of Science for Open and Environmental Systems, Graduate School of Science and Technology, Keio University, Kanagawa, Japan（信息与计算机科学中心，开放与环境系统科学学院，科技研究生学校，东京大学，神奈川，日本）；Bytedance Inc., China（字节跳动公司，中国）；Tsinghua University, Beijing, China（清华大学，北京，中国）

Zichun Wang, Hairong Shi, Bingzheng Wei, Yan Xu, Zihua Wang

阅读 PDF

arXiv 页面 ↗HTML 原文 ↗

AI总结：

提出MedVol-R1框架，通过强化学习将临床推理解耦为可验证的2D证据锚点，再传播为3D掩膜，实现体积推理分割，在多个基准上达到最优性能。

摘要：

体积推理分割（VRS）旨在根据自由形式的临床查询在3D医学扫描中分割目标区域，其中所指对象通常是隐含的，需要医学知识和体积基础推理。现有方法通常依赖专门的分割标记将语言与掩膜解码连接起来，但这种耦合将决策过程压缩为不透明的潜在表示，限制了可解释性和对多样化叙述表达的泛化能力。在本文中，我们提出MedVol-R1，一种基于强化学习的VRS框架，明确地将证据基础与体积描绘解耦：LVLM将临床推理定位到可验证的2D证据锚点（关键轴向切片和2D边界框），然后由冻结的MedSAM2模块将其传播为连贯的3D掩膜。我们使用冷启动监督微调后接GRPO来训练MedVol-R1，并由多组件奖励引导，该奖励鼓励信息性证据选择、准确的2D空间定位和跨切片体积连贯性，无需昂贵的思维链注释。在M3D-Seg基准的CT-ORG、AbdomenCT-1K和KiTS23上的实验表明，MedVol-R1一致优于强基线并达到最先进性能，强化学习相比纯监督微调提供了明显增益。

英文摘要：

Volumetric Reasoning Segmentation (VRS) aims to segment a target region in a 3D medical scan from a free-form clinical query, where the referent is often implicit and requires both medical knowledge and volume-grounded reasoning. Existing methods typically rely on specialized segmentation tokens to connect language with mask decoding, but this coupling collapses the decision process into opaque latent representations, limiting interpretability and generalization to diverse narrative expressions. In this paper, we present MedVol-R1, a reinforcement learning-based framework for VRS that explicitly decouples evidence grounding from volumetric delineation: the LVLM grounds clinical reasoning to a verifiable 2D evidence anchor (key axial slice and 2D bounding boxes), which is then propagated into a coherent 3D mask by a frozen MedSAM2 module. We train MedVol-R1 with cold-start supervised fine-tuning followed by GRPO, guided by a multi-component reward that encourages informative evidence selection, accurate 2D spatial grounding, and cross-slice volumetric coherence, without requiring costly chain-of-thought annotations. Experiments on CT-ORG, AbdomenCT-1K, and KiTS23 from the M3D-Seg benchmark demonstrate that MedVol-R1 consistently outperforms strong baselines and achieves state-of-the-art performance, with reinforcement learning providing clear gains over pure supervised fine-tuning.

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。