The ACR Data Science Institute structures artificial intelligence development to optimize radiology care
The American College of Radiology Data Science Institute (DSI) will provide the framework, strategy and focus to move artificial intelligence (AI) from concept into everyday radiology practice.
“The ACR DSI’s open source framework for developing AI use cases for radiology defines standards for training, testing, validating, integrating, and monitoring AI algorithms in clinical practice. This provides the ACR DSI and other organizations a standardized platform for building AI use cases that will help optimize radiology practice and improve patient care,” said Bibb Allen, MD, FACR, chief medical officer of the ACR Data Science Institute.
ACR DSI end-to-end lifecycle will:
- Create an open-source standard framework available to medical organizations, institutions and developers for radiology AI-use case development
- Define specific ACR DSI AI use cases built around these standards that are the most relevant needs of the specialty
- Provide a standardized pathway for algorithm validation and certification that ensures algorithm effectiveness and patient safety and helps to expedite the FDA regulatory review processes
- Create radiology workflow interoperability standards and pathways for incorporating AI algorithms into the clinical workflow
- Provide ongoing post-market assessment of algorithm performance and effectiveness through its AI Registry
“Disparate AI development underway may not address the right clinical questions nor be compatible with multiple computer systems. The ACR DSI will organize these efforts to ensure that AI services are effective, improve patient care, and are compatible with modern electronic workflow and health record technologies,” said Keith Dreyer, DO, PHD, chief science officer of the ACR Data Science Institute.
Please visit the ACR Data Science Institute Booth (# 8547) in the Machine Learning Showcase (North Building, Hall A) at the 2017 Radiological Society of North American (RSNA) annual meeting.
