NVIDIA Partners With American College of Radiology To Drive AI Adoption In Diagnostic Radiology




NVIDIA Clara / Credit: NVIDIA

Earlier this week, NVIDIA and the American College of Radiology (ACR) announced a partnership that enables thousands of radiologists nationwide to create and use artificial intelligence for diagnostic radiology in their facilities while using their own data to meet their own clinical needs. This news comes after a successful three-month pilot program by both parties.

ACR is going to integrate the NVIDIA Clara AI toolkit into the ACR Data Science Institute ACR AI-LAB — which is a free software platform that will become available to more than 38,000 ACR members and radiology professionals to build, share, locally adapt, and validate AI algorithms while ensuring patient data stays protected.

The NVIDIA Clara AI toolkit is a part of the NVIDIA Clara developer platform, which was built to enable software-defined medical instruments and intelligent workflows. Plus NVIDIA Clara consists of libraries for data and image processing, AI model processing, and visualization. With AI, the toolkit includes libraries for data annotation, model training, model adaptation, model federation, and large-scale deployment.

“NVIDIA builds platforms that democratize the use of AI and we purpose-built the Clara AI toolkit to give every radiologist the opportunity to develop AI tools that are customized to their patients and their clinical practice,” said NVIDIA VP of Healthcare Kimberly Powell. “Our successful pilot with the ACR is the first of many that will make AI more accessible to the entire field of radiology.”

To make the vision of ACR AI-LAB a reality, it requires the collaboration of the whole ecosystem including industry leaders, healthcare startups, and leading research institutions. NVIDIA Clara is used to power GE Healthcare’s Edison AI platform and the Nuance AI Marketplace and both of these platforms are supporting the AI-LAB.

“This collaboration marks a significant milestone in an extraordinary ACR Data Science Institute project, helping enable the launch of the ACR AI-LAB, giving radiologists in any practice environment an opportunity to become involved in AI development at their own institutions, using their own patient data to meet their own clinical needs,” added Bibb Allen Jr., M.D., FACR — chief medical officer of the Data Science Institute at the American College of Radiology.

In order to determine the assets and pathways necessary to enable facilities to work together, an initial pilot was deployed with the Ohio State University (OSU) and the Massachusetts General Hospital and Brigham and Women’s Hospital’s Center for Clinical Data Science (CCDS). By bringing an AI model to the patient data rather than patient data to the model, it can help increase diversity in algorithm training, facilitate validation of the algorithms, and enable radiologists to learn the critical steps needed to adapt algorithms to their institutions’ clinical needs.

By using the NVIDIA Clara AI toolkit, OSU was able to quickly import a pre-trained model developed by CCDS. And the model was customized to local variables and successfully labeled OSU data for further testing and improvement of the algorithm — all of which happened behind their own firewall. And this resulted in a highly accurate and enhanced cardiac computed tomography angiography model. This shared approach reduced algorithm training, validation and testing times by days.

“This software will offer radiologists, without computer programming experience, the ability to build and improve AI algorithms without the need to share their data,” explained Dr. Keith J. Dreyer — chief data science officer at Partners Healthcare and associate professor of radiology at Harvard Medical School. “Algorithms typically work best within the sites where they were trained, but those limited training sets are not always representative of the population at large. Training AI models on data from diverse sites helps ensure resiliency while reducing algorithm bias, resulting in improved inference across broader populations.”

The NVIDIA Clara AI toolkit powered the architecture used in the pilot program and it enabled data aggregation, image annotation, image pre-processing/transformation, algorithm transfer, and local computing for algorithm improvement.

“Enabling a network of artificial intelligence between hospitals will create more robust algorithms, greater efficiencies and likely lead to better patient outcomes,” noted Richard D. White, MD, MS (FACR,FACC) — chair of the Department of Radiology and Medical Imaging Informatics at the Ohio State University Wexner Medical Center. “This will give us access to high-quality algorithms that will help us accelerate deep learning and machine learning in healthcare.”

Keith Bigelow, SVP of Edison Portfolio Strategy at GE Healthcare, pointed out that GE Healthcare can lower costs and improve patient outcomes by “accelerating the number of algorithms created and seamlessly deployed to Edison-powered healthcare devices and applications in hospitals nationwide.” This is made possible by GE’s support of the ACR community’s AI-LAB efforts and tapping into the power of NVIDIA’s Clara AI platform.

And Nuance’s vice president and general manager of Healthcare Diagnostics Karen Holzbergerz explained that combining the strength of the NVIDIA Clara AI platform with the scale of the Nuance AI Marketplace for Diagnostic Imaging will empower ACR AI-LAB developers to rapidly build and deploy AI algorithms into the existing workflows of over 70% of all radiologists across more than 5,800 connected healthcare facilities.

“Furthermore, the ubiquitous footprint of Nuance PowerScribe radiology reporting and PowerShare image-sharing solutions provides subscribers of our AI Marketplace with immediate access to the largest storefront of imaging AI algorithms that can be automatically integrated into the radiology reporting and interpretation tools they use every day,” Holzbergerz continued.

 






NVIDIA

Leave a Reply

Your email address will not be published.