DICOM Medical Imaging MCP Server

HealthcarePython

Query, read, and move data on DICOM medical imaging servers

Available Tools

query_patients

Search for patients based on criteria like name, ID, or birth date

patient_namepatient_idbirth_dateattribute_preset

query_studies

Find studies using patient ID, date, modality, description, accession number, or Study UID

patient_idstudy_datemodalitystudy_descriptionaccession_numberstudy_instance_uidattribute_preset

query_series

Locate series within a specific study using modality, series number/description, or Series UID

study_instance_uidmodalityseries_numberseries_descriptionseries_instance_uidattribute_preset

query_instances

Find individual instances (images/objects) within a series using instance number or SOP Instance UID

series_instance_uidinstance_numbersop_instance_uidattribute_preset

extract_pdf_text_from_dicom

Retrieve a specific DICOM instance containing an encapsulated PDF and extract its text content

sop_instance_uid

move_series

Send a specific DICOM series to another configured DICOM node using C-MOVE

series_instance_uiddestination_node

move_study

Send an entire DICOM study to another configured DICOM node using C-MOVE

study_instance_uiddestination_node

list_dicom_nodes

Show the currently active DICOM node and list all configured nodes

switch_dicom_node

Change the active DICOM node for subsequent operations

node_name

verify_connection

Test the DICOM network connection to the currently active node using C-ECHO

get_attribute_presets

List the available levels of detail (minimal, standard, extended) for metadata query results

The DICOM MCP server enables AI assistants to interact with medical imaging systems through the DICOM protocol. It provides tools to search for patients, studies, and series, extract text from PDF reports embedded in DICOM files, and move imaging data between DICOM nodes. This integration allows AI assistants to access and analyze medical imaging data while maintaining proper DICOM communication standards.

Overview

The DICOM MCP server creates a bridge between AI assistants and medical imaging systems that use the DICOM protocol (PACS, VNA, etc.). This integration enables AI systems to search for medical imaging data, extract clinical reports, and transfer images between systems.

Installation

You can install the DICOM MCP server using uv or pip:

uv tool install dicom-mcp

Or with pip:

pip install dicom-mcp

For development purposes, you can clone the repository:

git clone https://github.com/ChristianHinge/dicom-mcp
cd dicom-mcp
uv venv
source .venv/bin/activate
uv pip install -e ".[dev]"

Configuration

The DICOM MCP server requires a YAML configuration file that defines the DICOM nodes and calling AE titles. Create a file named config.yaml with the following structure:

nodes:
  main:
    host: "localhost"
    port: 4242
    ae_title: "ORTHANC"
    description: "Local Orthanc DICOM server"
  
  # Add additional nodes as needed
  ai_node:
    host: "ai-server.example.com"
    port: 11112
    ae_title: "AI_PROCESSOR"
    description: "AI processing node"

current_node: "main"
calling_aet: "MCPSCU"

Testing with a Sample DICOM Server

If you don't have access to a DICOM server, you can set up a local Orthanc server for testing:

Clone the repository and install test dependencies:
```
pip install -e ".[dev]"
```

Start the Orthanc Docker container:

cd tests
docker compose up -d
cd ..
pytest  # uploads dummy PDF data to the Orthanc server

Access the Orthanc UI at http://localhost:8042/

Usage Scenarios

The DICOM MCP server enables AI assistants to:

Search for patients, studies, and series using various criteria
Extract text from clinical reports stored as PDFs in DICOM objects
Move imaging data between DICOM nodes (e.g., sending to AI processing systems)
Verify connections to DICOM servers

Important Warning

DICOM-MCP is not intended for clinical use and should not be connected to live hospital databases or systems containing patient-sensitive data. Doing so could lead to data loss or leakage of patient information. For complete data privacy, consider using DICOM-MCP with locally hosted open-weight LLMs.

Example Workflow

A typical interaction might involve:

Querying for a patient by name
Finding relevant studies for that patient
Locating specific series within those studies
Extracting report text or moving images to another system for processing

The tools can be chained together to answer complex clinical questions while maintaining proper DICOM communication protocols.