An initiative that aims to demonstrate the feasibility of creating investigational artificial intelligence models from image data is taking another step forward in testing.
Radiologists from seven healthcare organizations will be using the ACR AI-LAB to further use of the approach to create models from image data without the use of a programming language.
In this next phase, an AI model developed at one institution can be evaluated and optimized at each of the seven organizations, enabling them to fine-tune a model for their own investigational use.
The project is significant because it seeks to demonstrate the feasibility of enabling organizations to develop high-quality algorithms that address local clinical needs. The ability to enable such optimization is critical to facilitating commercial use of algorithms.
The pilot, which originally included Massachusetts General Hospital and The Ohio State University, now also includes Lahey Hospital and Medical Center, Emory University, The University of Washington, the University of California San Francisco and Brigham and Women’s Hospital.
Brigham and Women’s Hospital
The American College of Radiology is playing a lead role in the research, offering AI-LAB, a free software platform that will be made available to its more than 38,0000 members and other radiology professionals. The combined capabilities will enable radiologists to build, share, locally adapt and validate AI algorithms.
ACR is collaborating with NVIDIA to extend the use of artificial intelligence for diagnostic radiology. Earlier this year, they reached an agreement to expand the number of radiologists that can use an NVIDIA application and data sharing toolkit for use in their own facilities with their own data to meet their specific clinical needs. NVIDIA is providing software and edge infrastructure, and Nuance is providing last-mile integration to the participating radiologist.
NVIDIA will provide its NVIDIA Clara AI software toolkits at no cost to the institutions to perform the annotation creation, transfer learning and pipeline integration. In addition, Nuance will provide the last-mile technology required to integrate AI for the participating radiologist. Once the pilot is complete, the initiative is anticipated to progressively expand to all institutions interested in participating.
“(This) marks a major step in accelerating the development of AI for medical imaging,” says Bibb Allen Jr., MD, chief medical officer of the ACR Data Science Institute. “We know algorithms can underperform when deployed at sites where they weren’t trained. Now, radiologists in the pilot program will have access to AI algorithms developed outside their institutions in order evaluate a model’s performance using their own data and, as necessary, retrain the algorithm using their local data to enhance its performance.”
Sharing local AI models from image data between institutions for fine tuning—while patient information remains securely on site at the originating institution—has not previously been done successfully in radiology at this scale. This is due, in part, to the variability in how medical images are created, including the equipment, software and protocols used.
Date: July 09, 2019
Source: Health Data Management