Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Multi-Agent Intelligence for Multidisciplinary Decision-Making in Gastrointestinal Oncology

Rongzhao Zhang, Junqiao Wang, Shuyun Yang, Mouxiao Bian, Chihao Zhang, Dongyang Wang, Qiujuan Yan, Yun Zhong, Yuwei Bai, Guanxu Zhu, Kangkun Mao, Miao Wang, Chao Ding, Renjie Lu, Lei Wang, Lei Zheng, Tao Zheng, Xi Wang, Zhuo Fan, Bing Han, Meiling Liu, Luyi Jiang, Dongming Shan, Wenzhong Jin, Jiwei Yu, Zheng Wang, Jie Xu, Meng Luo

TL;DR

<3-5 sentence high-level summary>GI oncology demands integration of endoscopic, radiologic, and laboratory data for personalized management. The authors present a hierarchical Multi-Agent Framework with modality-specific agents and a central MDT-Core that aggregates intermediate reasoning into final MDT reports, addressing context dilution and hallucination seen in monolithic MLLMs. Evaluated on 2,174 cases from multiple institutions, the framework achieves a composite expert score of 4.60/5, outperforming a single-agent baseline and improving safety and reasoning fidelity. Ablation studies confirm the critical roles of Radiology, Endoscopy, and explicit cross-modal conflict detection, supporting scalable, interpretable decision support aligned with real-world MDT workflows.

Abstract

Multimodal clinical reasoning in the field of gastrointestinal (GI) oncology necessitates the integrated interpretation of endoscopic imagery, radiological data, and biochemical markers. Despite the evident potential exhibited by Multimodal Large Language Models (MLLMs), they frequently encounter challenges such as context dilution and hallucination when confronted with intricate, heterogeneous medical histories. In order to address these limitations, a hierarchical Multi-Agent Framework is proposed, which emulates the collaborative workflow of a human Multidisciplinary Team (MDT). The system attained a composite expert evaluation score of 4.60/5.00, thereby demonstrating a substantial improvement over the monolithic baseline. It is noteworthy that the agent-based architecture yielded the most substantial enhancements in reasoning logic and medical accuracy. The findings indicate that mimetic, agent-based collaboration provides a scalable, interpretable, and clinically robust paradigm for automated decision support in oncology.

Multi-Agent Intelligence for Multidisciplinary Decision-Making in Gastrointestinal Oncology

TL;DR

<3-5 sentence high-level summary>GI oncology demands integration of endoscopic, radiologic, and laboratory data for personalized management. The authors present a hierarchical Multi-Agent Framework with modality-specific agents and a central MDT-Core that aggregates intermediate reasoning into final MDT reports, addressing context dilution and hallucination seen in monolithic MLLMs. Evaluated on 2,174 cases from multiple institutions, the framework achieves a composite expert score of 4.60/5, outperforming a single-agent baseline and improving safety and reasoning fidelity. Ablation studies confirm the critical roles of Radiology, Endoscopy, and explicit cross-modal conflict detection, supporting scalable, interpretable decision support aligned with real-world MDT workflows.

Abstract

Multimodal clinical reasoning in the field of gastrointestinal (GI) oncology necessitates the integrated interpretation of endoscopic imagery, radiological data, and biochemical markers. Despite the evident potential exhibited by Multimodal Large Language Models (MLLMs), they frequently encounter challenges such as context dilution and hallucination when confronted with intricate, heterogeneous medical histories. In order to address these limitations, a hierarchical Multi-Agent Framework is proposed, which emulates the collaborative workflow of a human Multidisciplinary Team (MDT). The system attained a composite expert evaluation score of 4.60/5.00, thereby demonstrating a substantial improvement over the monolithic baseline. It is noteworthy that the agent-based architecture yielded the most substantial enhancements in reasoning logic and medical accuracy. The findings indicate that mimetic, agent-based collaboration provides a scalable, interpretable, and clinically robust paradigm for automated decision support in oncology.

Paper Structure

This paper contains 27 sections, 1 figure, 3 tables, 1 algorithm.

Table of Contents

  1. Introduction
  2. Related Work
  3. Multimodal AI in Medical Diagnostics
  4. Clinical Decision Support via Large Language Models
  5. Methods
  6. Problem Formulation and Multi-Agent Architecture
  7. Training of Domain‑Specific Agents
  8. Text Agent
  9. Endoscopy Agent
  10. Radiology Agent
  11. Laboratory Agent
  12. MDT‑Core Agent
  13. SFT Data Curation
  14. Inference Workflow and Algorithms
  15. Dataset and Experimental Setup
  16. ...and 12 more sections

Figures (1)

  • Figure 1: The proposed Multi-Agent Architecture, where multiple agents process corresponding modal information and incorporate it into the evidence container. After cross-modal conflict detection, the data are processed by the MDT-Core agent to generate a standardized final report.