Mainecoon

Mainecoon is a powerful web-based digital pathology viewer designed for researchers, pathologists, and healthcare professionals to explore and analyze pathology images directly in the browser.

Live DEMO   

Demo Videos

• Video - Introduction
• Video - Rendering for large bulk annotations
• Video - Display the DICOM WG26 Connectathon 2024 Dataset

About

• Mainecoon features a user-friendly interface with full support for responsive web design (RWD), eliminating the need for additional software installation.
• Provides five annotation tools, including "point, polyline, polygon, ellipse, and rectangle"
• Validated the "Evidence Creator" and "Viewer" actors in the 2024 DICOM WG26 Connectathon, enabling the creating annotations, is compliant with DICOM standards, featuring Microscopy Bulk Simple Annotations (also called ANN)

Key Features

Network support

• Integration with any DICOMweb image archive, including Raccoon, Orthanc, and dcm4chee server
  • Retrieve methods: WADO-URL and WADO-RS
  • Store annotations: STOW-RS

Authentication: OAuth2

• OAuth2 - Enable the OAuth2 to modify the oauthConfig.json
  • Note: We used Keycloak to test the function of OAuth2.

Viewer

• Whole Slide Image (WSI) Viewer: Display WSIs using efficient tiling and progressive loading techniques.
• Annotation Support: Create and manage annotations of graphic types defined in the DICOM Microscopy Bulk Simple Annotations Module for precise image labeling and analysis. Annotations can be stored via STOW-RS on the connected DICOMweb server.
• Advanced Image Interaction: Pan, zoom, and rotate multi-layer images through OpenLayers for detailed analysis.

Installation

Before starting, ensure you have configured the environment variables as needed. See the DICOMweb Server Configuration section for more information.

Docker

To quickly start Mainecoon using Docker:

docker compose up -d

Build from source

Dependencies

To run Mainecoon from the source, ensure you have the following installed:

• Node.js (v18.18 or later)
• npm

Building

Install the dependencies and build the project:

npm install
npm run build

Running

node .\app\server.js

The viewer should now be accessible at http://localhost:3000.

DICOMweb Server Configuration

To set up the DICOMweb Server, make changes to the following file: .src/config/DICOMWebServer.config.js

DICOMweb Server Configuration Example

This configuration file is to configure the DICOMweb server details for different environments.

export default {
  SERVER_NAME: {
    QIDO: {
      enableHTTPS: true,
      hostname: "your.server.com",
      port: "",
      pathname: "/dicom-web",
      Token: null,
    },
    WADO: {
      enableHTTPS: true,
      hostname: "your.server.com",
      port: "",
      URI_pathname: "/dicom-web/wado",
      RS_pathname: "/dicom-web",
      Mode: "rs",
      Token: null,
    }
  }
}

Environment Variables

The following variables are used to configure the runtime settings:

• PORT: The port number the server listens on (default: 3000).

Roadmap

• Support for ICC Profile display to ensure accurate color representation.
• Support displaying JPEG2000 Lossless compression via WADO-RS option: image/jp2.
• Support displaying the comprehensive 3D DICOM SR.
• Support displaying additional DICOM annotation types including parametric maps, saliency maps, and segmentation.

Acknowledgement

• To acknowledge the Mainecoon in an academic publication, please cite

  • Hsu, CW., Yang, SW., Lee, YT. et al. Mainecoon: Implementing an Open-Source Web Viewer for DICOM Whole Slide Images with AI-Integrated PACS for Digital Pathology. J Digit Imaging. Inform. med. (2025). https://doi.org/10.1007/s10278-025-01425-6

• Grants from the Ministry of Science and Technology Taiwan supported this project.

• We acknowledge H99 teams at Taipei Veterans General Hospital (TVGH) and the Smile Lab at National Cheng Kung University for validation.

• Dr.Yi-Chen Yeh from the Department of Pathology and Laboratory Medicine TVGH, provides many useful suggestions in many aspects of the clinical domain.

License

This project is licensed under the MIT License.