The modern National Radiology Data Registry, announced this week by the American College of Radiology, is capable of monitoring AI results and collecting a range of contextual information such as patient data, clinical metadata and radiology report results.
It is intended to compare the performance of individual radiology practices with aggregated national performance results from other centers using identical or similar products, ACR says.
WHY IT’S IMPORTANT
Clinical centers and AI developers can employ ACR Rate AIwhich will monitor the performance of a range of artificial intelligence imaging algorithms in real-world clinical settings to provide reports on the performance of the deployed artificial intelligence, the organization said on Monday.
The ACR Data Science Institute will facilitate data governance, which could also aid developers improve future versions of algorithms, ACR says.
According to Dr. Christoph Wald, vice chair of the ACR Council of Chancellors and chair of the ACR Informatics Committee, legacy radiology systems were not built to support efforts to ensure the operation of algorithms used in clinical settings.
“Users of artificial intelligence technologies in radiological care must ensure that algorithms perform sufficiently in their local environment,” Wald said in a statement.
“With growing demand for imaging outpacing the supply of radiologists, artificial intelligence is seen as an essential tool to help fill the gap and enable radiologists to maintain high standards of care while meeting growing demands,” Dr. Woojin Kim, medical director of ACR DSI, added.
ACR DSI’s simplified surveillance offers “a tangible, real-world approach to the challenge radiologists face today,” Wald said.
Radiology facilities participating in the program will have access to:
- Monitor algorithm stability over time, including imaging hardware, protocols, and software versions.
- Constant supervision of the compliance/discrepancy of AI results with radiology reports.
- Aggregated observations in reports and dashboards.
A BIGGER TREND
Innovation in radiology is advancing rapidly, and many clinical and financial applications of AI have demonstrated its value, but the long-term safety of algorithms remains an vital consideration for ACR.
Scottsdale, Arizona-based SimonMed is one of many practices where AI tools have improved workflow. Algorithms have been found to change radiology reports approximately 82% faster than readings without automation.
“When it comes to improved diagnosis, the tools can be truly extraordinary, so it’s important to stay open and curious because this is a rapidly evolving field,” Dr. John Simon, CEO of SimonMed, said in February.
Earlier in July, ACR launched the Recognized Center for Healthcare as a first-of-its-kind quality assurance program designed to benchmark radiology facilities that employ artificial intelligence in their imaging processes.
“Even an AI product approved by the U.S. Food and Drug Administration must be tested locally to ensure it works safely and as intended,” Wald cautioned in a statement from the quality assurance center.
“Practice leaders must implement safeguards to maximize the benefits of AI products while minimizing risks.”
ON RECORDING
“Assess-AI will play a key role in safely and effectively accelerating the clinical adoption of artificial intelligence in radiology by ensuring AI products perform optimally in clinical settings, so radiologists can focus on what matters most – providing high-quality care to their patients.” Kim said in a statement.
“The launch of Assess-AI, expanding its commitment to the radiology profession, is ACR’s latest step in helping radiologists implement artificial intelligence safely, effectively and transparently.”