الملخص
While medical multimodal large language models (Med-MLLMs) advance medical visual question answering (VQA), existing clinical workflow-inspired multi-agent frameworks suffer from interaction patterns and excessive computational overhead caused by redundant communication topologies. In this paper, we propose MedPrune, an efficient medical multimodal multi-agent collaboration framework that dynamically prunes both nodes and edges from the communication topology to enhance reasoning ability and token efficiency. Specifically, we first formulate the diagnostic process as a heterogeneous communication graph, where nodes represent specialist agents from various departments and edges capture intra- and inter-departmental interactions. Building on this graph, we introduce two sparsification mechanisms to enable adaptive collaborative evolution: (1) Heterogeneous Node Sparsification, which eliminates task-irrelevant specialist agents irrelevant to the current multimodal question via reinforcement learning-driven topological optimization, and (2) Heterogeneous Edge Sparsification, which selectively retains only the most diagnostically salient intra- and inter-departmental connections by jointly optimizing task performance and topological complexity. Extensive medical VQA experiments under full-set and few-shot training settings prove MedPrune surpasses multi-agent baselines and boosts token efficiency with strong adversarial robustness.
الكلمات المفتاحية
الموضوع
بيانات النشر
- المجلة
- غير متاح
- وصول مفتوح
- وصول مفتوح أخضر
اقتبس هذه المقالة
APA 7
Wan, J., Li, R., Chen, C., Hu, T., Yu, D., Jing, Y., Zhang, T., & Hong, R. (2026). MedPrune: Topology-Efficient Multimodal Multi-Agent Communication Evolution for Medical VQA Tasks. https://omanscience.com/ar/articles/medprune-topology-efficient-multimodal-multi-agent-communication-evolution-for-medical-vqa-tasks
MLA 9
Wan, Jiuheng, et al. "MedPrune: Topology-Efficient Multimodal Multi-Agent Communication Evolution for Medical VQA Tasks." https://omanscience.com/ar/articles/medprune-topology-efficient-multimodal-multi-agent-communication-evolution-for-medical-vqa-tasks.
شيكاغو (المؤلف–التاريخ)
Wan, Jiuheng, Runze Li, Chen Chen, Tingyuan Hu, Daiyang Yu, Yimin Jing, Taolin Zhang, and Richang Hong. 2026. "MedPrune: Topology-Efficient Multimodal Multi-Agent Communication Evolution for Medical VQA Tasks." https://omanscience.com/ar/articles/medprune-topology-efficient-multimodal-multi-agent-communication-evolution-for-medical-vqa-tasks.
هارفارد
Wan, J., Li, R., Chen, C., Hu, T., Yu, D., Jing, Y., Zhang, T. and Hong, R. (2026) 'MedPrune: Topology-Efficient Multimodal Multi-Agent Communication Evolution for Medical VQA Tasks', Available at: https://omanscience.com/ar/articles/medprune-topology-efficient-multimodal-multi-agent-communication-evolution-for-medical-vqa-tasks.
فانكوفر
Wan J, Li R, Chen C, Hu T, Yu D, Jing Y, et al. MedPrune: Topology-Efficient Multimodal Multi-Agent Communication Evolution for Medical VQA Tasks. https://omanscience.com/ar/articles/medprune-topology-efficient-multimodal-multi-agent-communication-evolution-for-medical-vqa-tasks
IEEE
J. Wan, R. Li, C. Chen, T. Hu, D. Yu, Y. Jing, T. Zhang, and R. Hong, "MedPrune: Topology-Efficient Multimodal Multi-Agent Communication Evolution for Medical VQA Tasks," https://omanscience.com/ar/articles/medprune-topology-efficient-multimodal-multi-agent-communication-evolution-for-medical-vqa-tasks.