mcp-registry/dicom-mcp

    ==================
      
       /// MCP ///
      /// DIC ///
        
    ==================
        
    [server:online]
    [protocol:ready]

dicom-mcp

by ChristianHinge

Model Context Protocol (MCP) server that lets LLM agents query, read and move data on DICOM servers (PACS / VNA).

56
18
Specialized
GitHub

Installation

1. Prerequisites
• Python 3.9 or newer
• Git
• Access to a DICOM-compatible server (PACS, Orthanc, etc.) and its connection parameters (AE-Title, Host, Port, Username/Password or TLS certificates if required).
2. Clone the repository
git clone https://github.com/ChristianHinge/dicom-mcp.git cd dicom-mcp
3. Create and activate a virtual environment (recommended)
python -m venv .venv source .venv/bin/activate # Windows: .venv\Scripts\activate
4. Install package dependencies
pip install -r requirements.txt
If no requirements.txt is present, install directly from the project root:
pip install -e .
5. Configuration
Create an environment file or export variables so the MCP layer can reach the DICOM server:
export DICOM_MCP_AET=MY_AET export DICOM_MCP_PEER_AET=PEER_AET export DICOM_MCP_PEER_HOST=127.0.0.1 export DICOM_MCP_PEER_PORT=104 export DICOM_MCP_USERNAME=xxxx # if your PACS needs auth export DICOM_MCP_PASSWORD=xxxx
Alternatively edit a provided `config.yaml` / `.env` template if it exists.
6. Run the MCP service (examples)
# Launch a local HTTP or gRPC MCP server (example entry-point) python -m dicom_mcp.server --host 0.0.0.0 --port 8080
7. Verify the installation
curl http://localhost:8080/health
or run the test suite:
pytest

Documentation

License: MIT License
Updated 7/30/2025