Skip to content

DICOM Connector Overview

Introduction

The Flywheel DICOM Connector automatically uploads medical images from Picture Archiving and Communication Systems (PACS) or scanners directly to your Flywheel instance via HTTPS. This service streamlines data acquisition by eliminating manual upload steps and ensuring consistent data organization.

Key Features

The DICOM Connector provides:

  • Automatic Data Transfer: Continuously monitors and uploads new DICOM data
  • Flexible Setup Options: Supports both PULL and PUSH configurations
  • Smart Organization: Uses routing strings to automatically organize data into the Flywheel hierarchy
  • Metadata Extraction: Extracts and maps DICOM metadata to Flywheel fields
  • Upload Control: Configurable opt-in or opt-out policies per scanner or study
  • Anonymization: Remove or obfuscates sensitive metadata following rules configured via de-identification profiles

Deployment Options

Flywheel manages all aspects of DICOM Connector deployment, configuration, and maintenance. The DICOM Connector can be deployed in two configurations:

  • Edge Device Deployment: Traditional on-premises approach where the Connector runs on customer-provided infrastructure (VM or hardware). DICOM communication stays on your controlled network, while only outbound HTTPS traffic reaches Flywheel Core. Required for PULL connectors; supports both PULL and PUSH.

  • In-Cluster Deployment: Newer option where the Connector runs as a service within Flywheel infrastructure. Uses mutual TLS (mTLS) to secure DICOM communication over the public internet. Reduces deployment complexity and costs. Supports PUSH connectors only.

Flywheel works with you to determine the most appropriate deployment model based on connector type requirements, security needs, infrastructure availability, and operational preferences.

See the Deployment Planning Guide and Infrastructure Reference for detailed guidance.

Connector Types

The DICOM Connector supports two operational modes:

Aspect PULL Connector PUSH Connector
Communication Direction Connector → Scanner Scanner → Connector
Who Initiates Connector queries scanner Scanner sends to Connector
DICOM Operations FIND-SCU, MOVE-SCU C-STORE (receive)
Deployment Options Edge device only Edge device or in-cluster
Best For Direct scanner access PACS integrations

Static IP Addresses Required

All DICOM devices require static IP addresses. The DIMSE protocol uses IP addresses as a primary authentication mechanism, and dynamic IP addresses create security risks and reliability issues.

Flywheel determines the appropriate connector type based on your network architecture and DICOM infrastructure.

See the Infrastructure Reference for detailed connector type specifications and network requirements.

How It Works

The DICOM Connector continuously monitors for new data (PULL queries scanners; PUSH receives from scanners). It extracts metadata, applies optional anonymization, and uploads complete series to Flywheel using routing strings for placement.

Upload policies control when data is transferred: opt-out (upload by default unless user opts out) or opt-in (upload only when user provides routing string).

See the Metadata Extraction Guide for detailed workflow information.

Documentation Structure

Select the appropriate guide based on your role and specific needs:

For Deployment Planning

Deployment Planning Guide helps you understand deployment options and prepare for DICOM Connector implementation.

For Scanner Operators and Researchers

User Guide provides step-by-step instructions for configuring scanners and routing data to correct Flywheel projects.

For Site Administrators

Admin Guide covers DICOM peer setup, upload policy configuration, and advanced system management.

For Technical Teams

Infrastructure Reference provides deployment architectures, networking requirements, and infrastructure specifications.

Hierarchy Mapping Guide explains how DICOM structures map to the Flywheel hierarchy and addresses the UID-based vs label-based identification challenge.

Metadata Extraction Guide covers DICOM metadata field extraction and processing mechanics.

Advanced Configuration Guide provides advanced metadata extraction, custom routing patterns, and acquisition naming options.

Getting Started

The basic workflow: identify target Group and Project, configure scanner with routing string (format: fw://group/project/subject/session), begin scanning, and verify data arrives in Flywheel.

The Connector organizes data within the Flywheel hierarchy (Group → Project → Subject → Session → Acquisition). When routing information is incomplete or missing, data uploads to default locations (Unknown Group or Unsorted Project) where administrators can manually organize it later.

See the Hierarchy Mapping Guide for details on how DICOM structures map to Flywheel containers.

Support Resources

For additional assistance:

  • Contact your institution's Flywheel Site Administrator for site-specific configuration details
  • Review troubleshooting sections in the role-specific guides above
  • Consult the DICOM conformance statement for technical specifications