MARCH: Multi-Agent Radiology Clinical Hierarchy for CT Report Generation

2026-04-17

AI Summary

This work addresses the clinical hallucination issues prevalent in existing automated 3D medical imaging report generation methods and their lack of an iterative validation mechanism inherent to radiological practice. To bridge this gap, the authors propose the first multi-agent framework that emulates the hierarchical structure of radiology departments, integrating three distinct roles—resident, fellow, and attending radiologist—to collaboratively draft reports, perform retrieval-augmented revision, and reach stance-driven consensus. The approach synergistically combines multi-scale CT feature extraction, retrieval-augmented generation, and a multi-agent dialogue consensus algorithm. Evaluated on the RadGenome-ChestCT dataset, the method significantly outperforms current state-of-the-art approaches, achieving superior performance in both clinical accuracy and linguistic quality metrics.

Technology Category

• Multiagent Systems: Modeling other Agents
• Cognitive Modeling & Cognitive Systems: Agent Architectures
• Computer Vision: Multi-modal Vision

Application Category

• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions

Abstract

Automated 3D radiology report generation often suffers from clinical hallucinations and a lack of the iterative verification found in human practice. While recent Vision-Language Models (VLMs) have advanced the field, they typically operate as monolithic "black-box" systems without the collaborative oversight characteristic of clinical workflows. To address these challenges, we propose MARCH (Multi-Agent Radiology Clinical Hierarchy), a multi-agent framework that emulates the professional hierarchy of radiology departments and assigns specialized roles to distinct agents. MARCH utilizes a Resident Agent for initial drafting with multi-scale CT feature extraction, multiple Fellow Agents for retrieval-augmented revision, and an Attending Agent that orchestrates an iterative, stance-based consensus discourse to resolve diagnostic discrepancies. On the RadGenome-ChestCT dataset, MARCH significantly outperforms state-of-the-art baselines in both clinical fidelity and linguistic accuracy. Our work demonstrates that modeling human-like organizational structures enhances the reliability of AI in high-stakes medical domains.

Problem

• clinical hallucinations
• radiology report generation
• iterative verification
• multi-agent system
• vision-language models

Innovation

• Multi-Agent System
• Radiology Report Generation
• Clinical Hierarchy
• Retrieval-Augmented Revision
• Iterative Consensus