MammoClaw: Towards Skill-Evolving Agent Harness for Breast Cancer Mammography Analysis

2026-09-25

AI Summary

This study addresses the lack of flexible and interpretable AI frameworks for mammographic analysis by proposing a training-free multimodal agent system. Centered on a frozen multimodal large language model, the method iteratively gathers evidence using lightweight medical imaging tools. It innovatively introduces a non-parametric skill evolution mechanism that transforms failure trajectories into reusable guidelines, enabling self-adaptation and transparent auditing of the agent. Experimental results demonstrate that this framework effectively optimizes tool-calling behaviors in BI-RADS assessment and breast density estimation tasks, significantly enhancing both diagnostic performance and interpretability.

Abstract

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.

Problem

• Mammography Analysis
• Breast Cancer
• Training-free Agent
• Skill Evolution
• Multimodal Large Language Models

Innovation

• Training-free Agent
• Skill Evolution
• Mammography Analysis
• MLLMs
• Non-parametric Adaptation