The ACR Recognized Center for Healthcare-AI, launched this past week by the American College of Radiology, is touted as a first-of-its-kind quality assurance program for radiology facilities seeking to make use of artificial intelligence in their imaging workflows.

WHY IT MATTERS
As radiology practices work toward and attest to the program's compliance goals, participation in ARCH-AI can help them implement safer and more effective AI products while helping radiologists provide better patient care, according to ACR – which lists recognition criteria such as establishing an interdisciplinary AI governance group and maintaining an inventory of AI algorithms with detailed documentation.

The program also prioritizes practices ensuring adherence to security and compliance measures; engaging in rigorous review and selection of algorithms; documenting use cases-focused training procedures; monitoring model performance for safety and efficacy; and contributing to the "Assess-AI" central AI registry for performance benchmarking.

ARCH-AI was built using established best practices, ACR leaders say, and is designed to offer expert consensus-based building blocks focused on infrastructure, processes and governance.

"ARCH-AI can help radiology practices structure QA processes that help them plan for what can go wrong, including the development of good AI governance practices, acceptance testing and effectiveness monitoring of AI products to ensure they continue to function as expected over time," said Dr. Keith J. Dreyer, chief science officer at ACR Data Science Institute and a longtime leader in AI-enabled radiology innovation.

ACR says practices that complete ARCH-AI attestation will receive a badge to display in their lobbies to demonstrate their commitment to AI safety to patients, payers and referring physicians.

THE LARGER TREND
Artificial intelligence and machine learning applications are already transforming how radiology is practiced, and they hold huge potential to help drive more efficient and higher-quality imaging processes as AI continues to evolve.

But, like any clinical application of AI, it's critically important to roll them out carefully and deliberately, with a keen eye toward safety and efficacy. That's what this and other frameworks are designed to help ensure.

ON THE RECORD
"AI is different from previous technologies," said Dr. Christoph Wald, vice chair of the ACR Board of Chancellors and chair of the ACR Commission on Informatics, in a statement. "Even a U.S. Food and Drug Administration-cleared AI product must be tested locally to ensure it works safely and as intended. Practice leaders must put safeguards in place to maximize the benefit of AI products while minimizing risk."

ARCH-AI, he said, offers "a low-cost, efficient system to help sites do that."

Mike Miliard is executive editor of Healthcare IT News
Email the writer: mike.miliard@himssmedia.com
Healthcare IT News is a HIMSS publication.

The HIMSS AI in Healthcare Forum is scheduled to take place Sept. 5-6 in Boston. Learn more and register.