Saved in:
| Main Authors: | , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2603.05860 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866911492371120128 |
|---|---|
| author | Fan, Lin Dai, Pengyu Deng, Zhipeng Wang, Haolin Gong, Xun Zheng, Yefeng Ou, Yafei |
| author_facet | Fan, Lin Dai, Pengyu Deng, Zhipeng Wang, Haolin Gong, Xun Zheng, Yefeng Ou, Yafei |
| contents | Clinical image interpretation is inherently multi-step and tool-centric: clinicians iteratively combine visual evidence with patient context, quantify findings, and refine their decisions through a sequence of specialized procedures. While LLM-based agents promise to orchestrate such heterogeneous medical tools, existing systems treat tool sets and invocation strategies as static after deployment. This design is brittle under real-world domain shifts, across tasks, and evolving diagnostic requirements, where predefined tool chains frequently degrade and demand costly manual re-design. We propose MACRO, a self-evolving, experience-augmented medical agent that shifts from static tool composition to experience-driven tool discovery. From verified execution trajectories, the agent autonomously identifies recurring effective multi-step tool sequences, synthesizes them into reusable composite tools, and registers these as new high-level primitives that continuously expand its behavioral repertoire. A lightweight image-feature memory grounds tool selection in a visual-clinical context, while a GRPO-like training loop reinforces reliable invocation of discovered composites, enabling closed-loop self-improvement with minimal supervision. Extensive experiments across diverse medical imaging datasets and tasks demonstrate that autonomous composite tool discovery consistently improves multi-step orchestration accuracy and cross-domain generalization over strong baselines and recent state-of-the-art agentic methods, bridging the gap between brittle static tool use and adaptive, context-aware clinical AI assistance. Code will be available upon acceptance. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2603_05860 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Evolving Medical Imaging Agents via Experience-driven Self-skill Discovery Fan, Lin Dai, Pengyu Deng, Zhipeng Wang, Haolin Gong, Xun Zheng, Yefeng Ou, Yafei Artificial Intelligence Computer Vision and Pattern Recognition I.2.6; I.4.10; J.3 Clinical image interpretation is inherently multi-step and tool-centric: clinicians iteratively combine visual evidence with patient context, quantify findings, and refine their decisions through a sequence of specialized procedures. While LLM-based agents promise to orchestrate such heterogeneous medical tools, existing systems treat tool sets and invocation strategies as static after deployment. This design is brittle under real-world domain shifts, across tasks, and evolving diagnostic requirements, where predefined tool chains frequently degrade and demand costly manual re-design. We propose MACRO, a self-evolving, experience-augmented medical agent that shifts from static tool composition to experience-driven tool discovery. From verified execution trajectories, the agent autonomously identifies recurring effective multi-step tool sequences, synthesizes them into reusable composite tools, and registers these as new high-level primitives that continuously expand its behavioral repertoire. A lightweight image-feature memory grounds tool selection in a visual-clinical context, while a GRPO-like training loop reinforces reliable invocation of discovered composites, enabling closed-loop self-improvement with minimal supervision. Extensive experiments across diverse medical imaging datasets and tasks demonstrate that autonomous composite tool discovery consistently improves multi-step orchestration accuracy and cross-domain generalization over strong baselines and recent state-of-the-art agentic methods, bridging the gap between brittle static tool use and adaptive, context-aware clinical AI assistance. Code will be available upon acceptance. |
| title | Evolving Medical Imaging Agents via Experience-driven Self-skill Discovery |
| topic | Artificial Intelligence Computer Vision and Pattern Recognition I.2.6; I.4.10; J.3 |
| url | https://arxiv.org/abs/2603.05860 |