Our Panel
Steven Blumer, MD, MBA
Americas Director of Radiology Digital Medical Affairs Bayer Radiology
Joseph Cavallo, MD, MBA
Assistant Professor, Radiology and Biomedical Imaging Yale School of Medicine
Saurabh Jha, MBBS, MS
Associate Professor Radiology Penn Medicine, University of Pennsylvania Health System
Eliot Siegel, MD, FSIIM, FACR
Chief, Imaging Services University of Maryland School of Medicine
Radiology is at the forefront of harnessing the potential of Artificial Intelligence (AI). With hundreds of FDA-approved algorithms, how can Radiology Departments advocate for and deploy AI Imaging technology with operational efficiency? Bayer in Radiology hosted an Industry Connect Session at the Society for Imaging Informatics in Medicine (SIIM) 2024 Annual Meeting to discuss these issues with a panel of experts.
Dr. Steven Blumer, Bayer’s Americas Director of Radiology Digital Medical Affairs and practicing radiologist, led the robust conversation about the strategic process of adopting and deploying AI Solutions applications and the advantages of utilizing a platform over stand-alone apps, shedding light on the potential for enhanced operational efficiencies and increased ROI.
Panelists expressed a strong interest in Radiology AI adoption and shared a variety of use cases within their universities and health systems, including helping to manage efficiencies across the radiology workflow, triage apps to flag suspected pathologies and findings that warrant urgent review and treatment, detection apps to improve lesion detection, and quantification apps for automation of routine tasks.
Advantages of AI Adoption
Panelists discussed the use of AI and how it may affect the efficiency of radiologists. Saurabh Jha, MBBS, MS, Associate Professor of Radiology, Penn Medicine, University of Pennsylvania Health Systems, said, “AI is going to allow us to read more complex cases as well as read more cases in a lower amount of time.” He added that while outcomes of AI can be difficult to measure at the patient level so far; it is not difficult to measure the value of AI at the radiologist level. He provided several examples, including the advantage of AI in turnaround times for ER studies. He also noted the use of AI is reducing time to review CT scans for lung nodules. “The ROI that comes from this budding AI revolution is making us more efficient radiologists,” Dr. Jha said.
Navigating the Process of AI Imaging Adoption and Deployment
Dr. Blumer asked the experts how they were advocating for the adoption of AI in their radiology departments. Discussion centered on the strategic considerations and practical steps involved in adopting and deploying AI applications, addressing challenges and best practices for seamless integration.
As an early adopter of AI, Joseph Cavallo, MD, MBA, Assistant Professor of Radiology and Biomedical Imaging, Yale School of Medicine, said “It is very helpful to involve both the hospital leadership and interdisciplinary members from other sections outside of Radiology in the decision-making process and governance. While the majority of AI applications are within the radiology field right now, I think that is that number is going to change. Radiology may be the majority holder today, but cardiology, ophthalmology, pathology, lots of other imaging-based departments or imaging heavy departments are going to want AI applications,” Dr. Cavallo said.
“One of the benefits of being an early adopter as a Radiology Department is the opportunity to help construct governance. It gives you insight, oversight and ownership potentially of the Healthcare AI system or process, so you can make sure that you are a stakeholder and have appropriate input within the greater hospital enterprise,” Dr. Cavallo added.
Leveraging Platforms for Comprehensive Solutions
The panelists also discussed the advantages of utilizing a platform for AI deployment in radiology, emphasizing the potential for interoperability, scalability, and comprehensive management of AI applications, and enhanced ROI and operational efficiencies in contrast to stand-alone deployments.
Dr. Cavallo suggested that platforms offer benefits upfront. “If you find a platform that you like, you can add Radiology AI modules either from that platform or license through that platform and avoid the necessity of going through lengthy security reviews, IT integrations, and contracting.”
Eliott Siegel, MD, FSIIM, FACR, Chief, Imaging Services, University of Maryland School of Medicine, agreed and noted that platforms are an efficient way to scale AI Solutions applications. He said, “Platforms allow us the capability to try out algorithms relatively easily and quickly. Platforms also allow us to be able to become more aware of the algorithms that are out there. Finally, platforms provide a unified mechanism to be able to consume multiple different AI algorithms without having to learn multiple different AI systems.”
Looking to the future, the panelists agreed that radiologists who use AI Solutions will replace radiologists who do not. They also discussed the importance of training residents on Radiology AI, so they are prepared for the changes they are going to encounter. There is great excitement for the next ten years of AI advancements and the ways AI applications will help meet the growing demands of imaging and work to improve patient care.
October 2024
