Artificial intelligence (AI) is being used in medical practice to refine processes, streamline workflows, and translate myriad healthcare practitioner complexities into improved efficiencies. While the technology is still relatively young, and its applications still evolving, it has become more of a part of standard care and is offering healthcare practitioners and facilities increased scope for improvement and operational control.
There is a growing trend towards AI becoming a part of the standard of care, but the real question is how can radiologists drive AI implementation and how can it help them to potentially differentiate themselves within the organization.
Radiologists as early AI adopters
At the virtual SIIM 2021 Annual Meeting, Nina Kottler, MD, MS, Associate Chief Medical Officer, Clinical AI at Radiology Partners, unpacked the role of AI in healthcare and how radiologists can take a stronger role as an early adopter of AI to differentiate their own roles within the healthcare organization.
“I think that one of the most important things that radiologists have to learn is how AI can potentially improve quality and efficiency, and how it can change the bigger picture for radiologists as a whole,” she says. “The most important thing is to have a business case. If you don’t you won’t have AI – there aren’t extra dollars sitting around in healthcare waiting to be spent. We need to ensure that whatever algorithm model is being deployed has a proven business case.”
Building the business case for AI
This business case is to drive value, and this value sits in the eye of the beholder – the clinician and the company. It has to improve patient care and outcomes, but this is not the only metric. After all, improved outcomes is an expectation of healthcare, so it’s taking this further. As Kottler points out, there is movement towards investment into innovative technologies and “putting dollars towards improved outcomes”.
“I think that as we start transitioning from fee for service to fee for value, we’ll see a better business case for AI because there is absolutely value in the use of AI,” she explains. “While we’re not at the fee for value level yet, we have to manage both ends of the spectrum and think about the financial value of AI. This then depends on the stakeholder. If you’re talking to me in a radiology practice, it’s improved efficiencies which adds to increased recruitment or more services sold.”
If the conversation shifts to other stakeholders, such as the hospital or healthcare system, then yes, everyone cares about improved outcomes, but they are equally interested in value. In patients coming back to use other services, ensuring proper follow-ups, managing better screenings or identifying chronic conditions earlier on. With increased speed to care in areas such as stroke imaging, then not only are patient outcomes improved, but also cost to healthcare outcomes.
“The things that are going to be paid for are overall decreased patient care costs, direct savings from the AI investment into reducing follow-ups that aren’t necessary or catching conditions early on,” says Kottler. “Some research has found that these simple steps can save up to $300 million across the US which is significant.”
As the area moves into prediction algorithms that can determine who is more at risk, allowing for improved patient care for those patients early on before they have complications, this mitigates a very expensive problem for the sector. If it’s possible to predict and plan, then there is value in the AI, in the algorithms and in the cost.
“If you’re looking at the healthcare system, then the ROI changes as now it’s looking at length of state reduction throughout and reduced numbers of procedures that have significant implications downstream,” says Kottler. “To really capture the full value of AI, radiologists need to see a much bigger picture. It’s also about how having better radiology services impacts downstream. That’s why we talk about the quadruple aim of driving value – decreased costs, improved patient experiences, improved physician experiences and improved hospital and client experiences.”
Radiologists juggle multiple clients so they have to consider all of these metrics and how technology can potentially shift these into greater efficiencies and value. This makes AI, and the evolution of the technology around AI, unusual and exciting. As the technology starts to take on a broader role, it can support physicians in taking a more dominant role in leading the technology evolution.
“Whether this is assessing an AI solution, or deploying one, physicians are instrumental in ensuring that the clinical tools and implications are addressed,” says Kottler. “You need physicians to ensure that the quadruple value proposition is being driven forward, and I believe that radiology is in a perfect position to take on this role.”
Radiology’s role at the forefront of AI adoption across the health system
For Kottler, the radiologist has already been more immersed in AI developments with many sitting at the forefront of early adoption. The role sits at the center of medical care, seeing every patient across the gamut of discipline and healthcare profession, and providing continuous oversight. She believes that the radiologist is already in a very technical group having spent careers understanding how to interpret and use new technology. Radiologists have rediscovered innovation and technology many times, and AI is no different.
“I think the radiologist is in a really good position to bring in this new and evolving technology to clinical practice in a way that’s going to benefit patient care,” she concludes. “However, we need to be drivers. We need to learn more about it, to move beyond the hype or away from being paralyzed because when we recognize the value and get invested, we are going to be the drivers of this in the future.”
Kottler doesn’t believe that the radiologist will ever be replaced, but that the role will change and evolve alongside technology. As leaders of innovation, technology and information management, radiologists are poised to get so much from AI, to drive better value, to improve outcomes from data, and add context to the transition to a more evolved, AI-driven future.