arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2603.01115cs.CV
收藏

GuiDINO:重新思考医学图像分割中的视觉基础模型

LUMOS: Latent Universal Medical Priors for Segmentation

The University of Sydney, Sydney. NSW 2006, Australia Biological System \& Engineering Division, Lawrence Berkeley National Laboratory, Berkeley. CA 94720, USA Berkeley Biomedical Data Science Center, Lawrence Berkeley National Laboratory, Berkeley. CA 94720, USA

Zhuonan Liang, Wei Guo, Jie Gan, Yaxuan Song, Runnan Chen, Hang Chang, Weidong Cai

AI总结:

GuiDINO通过引导掩码机制提升医学图像分割性能,提供了一种参数高效的基础模型适应方法。

摘要:

视觉基础模型在医学图像分析中正被越来越多地采用。由于领域偏移,这些预训练模型在未完全微调或轻度适应的情况下,与医学图像分割需求不匹配。我们介绍了GuiDINO,一个框架,将原生基础模型重新定位为下游分割任务的视觉引导生成器。GuiDINO从DINOv3中提取视觉特征表示,并通过轻量级的TokenBook机制将其转换为空间引导掩码,该机制聚合了token-原型相似性。此引导掩码在多个分割骨干网络中门控特征激活,从而在保持医学专用架构的归纳偏置和效率的同时注入基础模型先验。训练依赖于一个引导监督目标损失,将引导掩码对齐到真实区域,可选地通过聚焦边界的hinge损失来增强细结构。GuiDINO还通过LoRA在DINOv3引导骨干上支持参数高效适应。在多样化的医学数据集和nnUNet式推理中,GuiDINO始终提高分割质量和边界鲁棒性,表明了微调的实用替代方案,并提供了一种新的视角,即基础模型如何最好地服务于医学视觉。

英文摘要:

General vision foundation models (VFMs) have been primarily developed on natural images, and their utility for medical image segmentation is therefore often considered to depend on costly adaptation or domain-specific fine-tuning. In this paper, we revisit this assumption from a different perspective: rather than requiring VFM segmentors to relearn visual regularities, we investigate whether the low-level visual priors necessary for anatomical delineation already lie dormant within general VFMs. We observe that frozen VFMs, despite lacking medical supervision, encode transferable visual regularities. These properties are not exclusive to natural images but are also fundamental to medical image understanding. Motivated by this observation, we propose Latent Universal Medical PriOrs for Segmentation (LUMOS), a novel framework that amplifies general VFM priors to conventional medical segmentors. LUMOS consists of two key components: (1) Pathfinder that distills visual cues from a frozen vision foundation model, and (2) Inspiror that sparks the conventional medical networks with spatial guidance from distilled visual regularities. In this way, the segmentor is relieved from learning complex visual regularities entirely from limited medical annotations and can instead focus on task-specific anatomical delineation. Across diverse medical datasets and token-based VFMs, LUMOS shows that general VFMs can serve as spatial prior generators when their frozen token spaces preserve patch-level pattern relevance. DINO provides stable matched-backbone gains, while SigLIP exposes VFM-specific sensitivity caused by its different token granularity and representation objective.

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。

↑