发表机构
Centre for Bioengineering, School of Engineering and Materials Science, Queen Mary University of London; Digital Environment Research Institute, Queen Mary University of London; UCL Hawkes Institute; Department of Medical Physics and Biomedical Engineering, University College London; Department of Biological and Biomedical Sciences, Yale University; Institute of Biomedical Engineering, Department of Engineering Science, University of Oxford（伦敦玛丽女王大学工程与材料科学学院生物工程中心; 伦敦玛丽女王大学数字环境研究所; 伦敦大学学院UCL霍克斯研究所；医学物理与生物医学工程系; 耶鲁大学生物与生物医学科学系; 牛津大学工程科学系生物医学工程研究所）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
研究少样本医学图像分割中支持集选择问题，提出强化学习框架，智能体直接预测最大化下游分割性能的支持集，实验表明该方法优于随机选择和现有方法，凸显支持集互补性及强化学习的潜力。
AI 中文摘要
少样本学习使医学图像分割模型仅用少量标记示例就能适应新任务。然而，适应性能很大程度上取决于为支持集选择哪些示例。有效的支持集应捕捉目标域内的相关变化并为适应提供信息，其组成样本应提供互补信息。尽管如此，现有的主动数据选择方法大多单独优先考虑样本，未明确考虑示例间的相互作用。在这项工作中，我们提出了一个用于少样本医学图像分割中支持集选择的强化学习框架，使支持集能联合优化而非通过独立样本评分。给定一组未标记的候选图像，智能体直接预测能最大化下游分割性能的支持集。在跨机构盆腔MRI数据集上的实验表明优于随机选择和当前最先进的方法。我们的发现凸显了支持集互补性对有效适应的重要性，并证明了强化学习在优化适应集方面的潜力。
英文摘要
Few-shot learning enables medical image segmentation models to adapt to new tasks using only a small number of labelled examples. However, adaptation performance depends strongly on which examples are selected for the support set. Effective support sets should capture relevant variation within the target domain and be informative for adaptation, with constituent samples providing complementary information. Despite this, existing active data selection approaches largely prioritise samples individually and do not explicitly account for interactions between examples. In this work, we propose a reinforcement learning framework for support-set selection in few-shot medical image segmentation, enabling support sets to be optimised jointly rather than through independent sample scoring. Given a pool of unlabelled candidate images, an agent directly predicts a support set that maximises downstream segmentation performance. Experiments on a cross-institutional pelvic MRI dataset demonstrate improvements over random selection and current state-of-the-art methods. Our findings highlight the importance of support-set complementarity for effective adaptation and demonstrate the potential of reinforcement learning for optimising adaptation sets.
CommentsAccepted at EMA4MICCAI 2026 - The 2nd MICCAI Workshop on Efficient Medical AI
