CT-Flow: Orchestrating CT Interpretation Workflow with Model Context Protocol Servers

2026-02-23

AI Summary

This work addresses the limitation of existing 3D CT analysis methods, which predominantly rely on static, single-pass inference and fail to emulate the dynamic, tool-assisted interpretation process employed by radiologists. To bridge this gap, we propose the first agent framework based on the Model Context Protocol (MCP), introducing a tool-aware dynamic workflow paradigm that decomposes complex natural language queries into executable sequences of multi-step tool invocations, enabling seamless integration with clinical tools. We further construct CT-FlowBench, the first instruction-tuning benchmark tailored for 3D CT tool usage, integrating large vision-language models, 3D segmentation, and radiomics. Experiments demonstrate that our approach achieves state-of-the-art performance on both CT-FlowBench and standard 3D visual question answering benchmarks, surpassing baseline methods by 41% in diagnostic accuracy and attaining a 95% success rate in automated tool invocation.

Technology Category

• Cognitive Modeling & Cognitive Systems: Agent Architectures
• Computer Vision: 3D Computer Vision
• Multiagent Systems: Modeling other Agents

Application Category

• Economics, Online Markets and Human Computation: Architectures and workflows that use LLMs for crowd work
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Responsible Web: Machine-in-the-loop, human agency and autonomy

Abstract

Recent advances in Large Vision-Language Models (LVLMs) have shown strong potential for multi-modal radiological reasoning, particularly in tasks like diagnostic visual question answering (VQA) and radiology report generation. However, most existing approaches for 3D CT analysis largely rely on static, single-pass inference. In practice, clinical interpretation is a dynamic, tool-mediated workflow where radiologists iteratively review slices and use measurement, radiomics, and segmentation tools to refine findings. To bridge this gap, we propose CT-Flow, an agentic framework designed for interoperable volumetric interpretation. By leveraging the Model Context Protocol (MCP), CT-Flow shifts from closed-box inference to an open, tool-aware paradigm. We curate CT-FlowBench, the first large-scale instruction-tuning benchmark tailored for 3D CT tool-use and multi-step reasoning. Built upon this, CT-Flow functions as a clinical orchestrator capable of decomposing complex natural language queries into automated tool-use sequences. Experimental evaluations on CT-FlowBench and standard 3D VQA datasets demonstrate that CT-Flow achieves state-of-the-art performance, surpassing baseline models by 41% in diagnostic accuracy and achieving a 95% success rate in autonomous tool invocation. This work provides a scalable foundation for integrating autonomous, agentic intelligence into real-world clinical radiology.

Problem

• 3D CT interpretation
• clinical workflow
• tool-mediated reasoning
• multi-step reasoning
• radiology automation

Innovation

• CT-Flow
• Model Context Protocol
• tool-aware reasoning
• 3D CT interpretation
• agentic workflow