Workflow-friendly data curation, image annotation, and project management for machine learning in Radiology. Fully PACS-integrated and ready to deploy locally and securely within your institution's network.
Organize your machine learning project & easily find studies using the PACS-integrated app. Bulk import, anonymize, and collaboratively annotate images, all within routine clinical workflow.
Physician software developers and informatics enthusiasts, affiliated with the Johns Hopkins Radiology AI Lab (RAIL) & Malone Center for Engineering in Healthcare.
GRAIL is a web-based application with an intuitive UI, built with Vue Javascript. GRAIL integrates with clinical PACS using standard protocols and is easy to deploy with Node.js or as a desktop app with Electron.
Images Curated to Date
Define your machine learning project and set restrictions, collaborators, and markup definitions.
Find the right studies to add to your project using DICOM query or report search (using third party API).
Bulk import studies using a variety of import tools. Studies are automatically anonymized using HIPAA compliant protocols.
Review and clean collaboratively curated studies.
Label and annotate curated images. Annotation can be performed in real time while dictating cases during clinical workflow.
Review and download your images in a format ready for machine learning.
© Copyright 2019