Google Cloud Healthcare

The Google Cloud Healthcare API is a powerful option for storing medical imaging data in the cloud.

An alternative to deploying your own PACS is to use a software-as-a-service provider such as Google Cloud. The Cloud Healthcare API promises to be a scalable, secure, cost effective image storage solution for those willing to store their data in the cloud. It offers an almost-entirely complete DICOMWeb API which requires tokens generated via the OAuth 2.0 Sign In flow. Images can even be transcoded on the fly if this is desired.

Setup a Google Cloud Healthcare Project

  • Create a Google Cloud account
  • Create a project in Google Cloud
  • Enable the Cloud Healthcare API for your project.
    • (Optional): Create a Dataset and Data Store for storing your DICOM data
  • Enable the Cloud Resource Manager API for your project.
    • Note: If you are having trouble finding the APIs, use the search box at the top of the Cloud console.
  • Go to APIs & Services > Credentials to create an OAuth Consent screen and fill in your application details.
    • Under Scopes for Google APIs, click "manually paste scopes".
    • Add the following scopes:
  • Go to APIs & Services > Credentials to create a new set of credentials:

    • Choose the "Web Application" type
    • Set up an OAuth 2.0 Client ID
    • Add your domain (e.g. http://localhost:3000) to Authorized JavaScript origins.
    • Add your domain, plus callback (e.g. http://localhost:3000/callback) to Authorized Redirect URIs.
    • Save your Client ID for later.
  • (Optional): Enable Public Datasets that are being hosted by Google:

Run the viewer with your OAuth Client ID

  1. Open the config/google.js file and change YOURCLIENTID to your Client ID value.
  2. Run the OHIF Viewer using the config/google.js configuration file
cd OHIFViewer
yarn install
APP_CONFIG=config/google.js yarn run dev

Running via Docker

The OHIF Viewer Docker container can be connected to Google Cloud Healthcare by providing a Client ID at runtime. This is a very simple method to get up and running.

  1. Install Docker (
  2. Run the Docker container, providing a Client ID as an environment variable. Client IDs look like
docker run --env CLIENT_ID=$CLIENT_ID --publish 5000:80 ohif/viewer:latest

results matching ""

    No results matching ""