March 15, 2021

AI Interoperability in Imaging White Paper Released for Public Comment

A new white paper on AI Interoperability in Imaging has been released for public comment by the Integrating the Healthcare Enterprise (IHE) Radiology Technical Committee. A joint effort between medical imaging societies, industry and the radiology community, the paper sets the landscape for artificial intelligence (AI) interoperability and function. It provides a comprehensive map of the needs, problems and challenges that must be addressed to achieve an ecosystem of interoperable products that support all the processes and tasks that make up AI in Imaging. An IHE video introduces the white paper’s content.

For emerging technologies like AI to be successful, they must be integrated into the medical imaging ecosystem. “IHE Profiles are ‘recipes’ for standards based workflow,” said Brian Bialecki, an American College of Radiology® (ACR®) representative on the committee. “IHE compliance assures consumers that products have been tested to be interoperable with other products — across multiple vendor actors — helping to establish their trustworthiness,” he said. The white paper suggests ways IHE Profiles developed in the future can be properly scoped to not overlook key issues in AI-related medical imaging.

The ACR has taken leadership across the broad spectrum of medical imaging AI, and many themes found in the IHE white paper directly correlate to ACR Data Science Institute® ( DSI) initiatives, from structured Define-AI use cases describing tasks for which AI can be applied, to the AI-LAB™ platform for managing datasets and repositories for model testing. ACR Assist® and ACR Common offer standardized annotations useful for AI. The ACR Model API targets AI model distribution and creates a lower bar to entry for developers. ACR Assess-AI and Certify-AI are tools for model validation and feedback. ACR Connect, the next generation TRIAD, encompasses de-identification and provides a platform for federated learning.

ACR members are encouraged to review the paper and provide feedback. Reviewers will gain a better understanding of how ACR initiatives are guiding the community in providing safe and effective solutions in the clinical adoption of promising AI innovations.