MammoClaw：面向乳腺癌钼靶影像分析的可进化技能智能体框架
MammoClaw: Towards Skill-Evolving Agent Harness for Breast Cancer Mammography Analysis
浏览论文内容
中文总结 AI 辅助
MammoClaw提出无需训练的智能体框架，利用冻结MLLM和钼靶工具进行BI-RADS评估与密度估计，通过技能进化提升性能，支持透明审计。
中文摘要 AI 辅助
在本工作中，我们探索了MammoClaw，一个无需训练的智能体框架，利用冻结的多模态大语言模型（MLLM）进行钼靶影像分析。为了支持智能体式研究，我们为智能体配备了轻量级的钼靶特异性工具，用于针对性图像分析，包括感兴趣区域（ROI）、配对视图和对侧乳房检查。MammoClaw通过这些工具迭代地收集证据，而技能进化通过将失败的轨迹转化为后续运行中可复用的指导，实现非参数化自适应。我们在BI-RADS评估和乳腺密度估计任务上评估了该框架。在我们的实验中，我们发现仅靠工具并不能可靠地提升性能，而在某些设置下，进化后的技能可以改善工具使用行为和性能。除了这些结果，MammoClaw还支持对证据获取、工具交互和失败模式的透明检查，便于分析和审计智能体行为。我们将这项工作视为对钼靶影像无训练、自进化智能体方法的探索性研究，并希望它能为此后关于钼靶特异性工具和自进化机制的研究提供一个具体起点。我们在以下网址发布了代码：https://this URL。
英文摘要
In this work, we explore MammoClaw, a training-free agent framework that leverages frozen MLLMs for mammography analysis. To support agentic investigation, we equip the agent with lightweight mammography-specific tools for targeted image analysis, including ROI, paired-view, and contralateral-breast examination. MammoClaw iteratively gathers evidence through these tools, while skill evolution enables non-parametric adaptation by transforming failed trajectories into reusable guidance for later runs. We evaluate the framework on BI-RADS assessment and breast density estimation tasks. In our experiments, we find that tools alone do not reliably improve performance, whereas evolved skills can improve tool-use behavior and performance in some settings. Beyond these results, MammoClaw enables transparent inspection of evidence acquisition, tool interactions, and failure modes, facilitating the analysis and auditing of agent behavior. We view this work as an exploratory study of training-free, self-evolving agentic approaches for mammography and hope it provides a concrete starting point for future work on mammography-specific tools and self-evolution mechanisms. We release our code at https://krishnakanthnakka.github.io/mammoclaw.
