This post was contributed by ACR Data Science Institute (DSI) Chief Medical Officer Bibb Allen Jr., MD, FACR.
The ACR Data Science Institute™ (ACR DSI) is working with other heavy hitters in the field to create a framework to move artificial intelligence (AI) into medical imaging workflow to improve patient care.
Just this week, the ACR DSI released its first use cases in the TOUCH-AI library to industry leaders for comment prior to the projected release of the use cases in the fall of 2018. Use cases are clinical scenarios in which AI use may improve care. Until now, there has been no national body to identify or develop such cases. No other national organizations are yet involved in determining if algorithms developed by vendors actually answer a relevant clinical question to improve care. This process is an initial focus of the ACR DSI.
The ACR-DSI will also co-sponsor an August 23–24 workshop on AI in Medical Imaging, along with the National Institute of Biomedical Imaging and Bioengineering (NIBIB) and others. The two-day event will discuss gaps in both foundational and translational research in machine learning, data needs for machine learning in clinical imaging, examples of machine learning in the imaging life cycle and implementation issues. The workshop proceedings will be published as a research roadmap to be shared with health professionals in academia, industry and government.
With each passing day, the ACR DSI is laying the groundwork to take artificial intelligence in medical imaging from a novel idea to an everyday experience. This is an exciting time for radiology. Be glad you are here.
- What are some use cases where you see that imaging AI can help radiologists do their jobs better (more efficiently)?
- Where do you think imaging AI will be in 5–10 years?
Please share your thoughts in the comment section below or share them on Engage (login required).