ChristianHinge/dicom-mcp

Overview

A Model Context Protocol (MCP) server that enables AI agents to query and retrieve medical images from DICOM servers (PACS). It wraps standard DICOM networking protocols into a tool-calling interface for LLMs.

Best for

Best for
Developers building AI assistants for radiology or medical imaging workflows

Use cases

• Query PACS for patient studies by ID or date range
• Retrieve DICOM images for automated analysis or reporting
• Integrate medical imaging workflows with AI assistants

How to use

Install

npx @modelcontextprotocol/inspector uv run dicom-mcp /path/to/your_config.yaml --transport stdio

Tools exposed

• query_patients
• query_studies
• query_series
• query_instances
• extract_pdf_text_from_dicom
• move_series
• move_study
• list_dicom_nodes
• switch_dicom_node
• verify_connection
• get_attribute_presets

Example client config

nodes:\n  main:\n    host: "localhost"\n    port: 4242 \n    ae_title: "ORTHANC"\n    description: "Local Orthanc DICOM server"\ncurrent_node: "main"\ncalling_aet: "MCPSCU"

Notes

A Model Context Protocol (MCP) server that enables AI agents to query and retrieve medical images from DICOM servers (PACS). It wraps standard DICOM networking protocols into a tool-calling interface for LLMs.

96 stars on GitHub. Last updated 2025-12-15. Licensed MIT.

Use cases

• Query PACS for patient studies by ID or date range
• Retrieve DICOM images for automated analysis or reporting
• Integrate medical imaging workflows with AI assistants

Pros

• Bridges medical imaging systems with LLM tool use
• Open source with active community (96 stars)
• Python-based, easy to extend or embed

Cons

• Requires access to a DICOM server (PACS) to be useful
• Limited to query/retrieve operations, no write or delete
• Niche use case outside healthcare IT

Indexed from awesome-mcp-servers-punkpeye and enriched against its public facts.