AI in Healthcare – The Effect on Current and Future Care

By: Ron DiGiaimo, MBA, FACHE, Orest Boyko, MD, PhD, John Montville, MBA, FACHE, FACMPE, FACCC, COA, and Bri Driggers

“AI” is the hottest buzzword for 2023 – and maybe all of the 2020s.  Some discussions of Artificial Intelligence (AI) have included additional wording at the start of the phrase such as “Assistive”, “Augmented”, “Autonomous”, or “Adoption”. Regardless of the viewpoint, AI is on everyone’s mind, and it’s got many people divided on their perception of it. As with many technological innovations and advancements, the healthcare community is excited by innovation if they are early adopters, or reticent on this innovation and willing to be late adopters who tend to be in the majority of providers.  

Some comments made in AI adoption discussions include “Is a robot going to take my job?” This tends to be the first question that late adopters ask when the topic of AI comes up in conversation.

Surely an industry full of highly skilled and specialized providers wouldn’t need to worry about their job security… right?

RCCS CEO Ron DiGiaimo and marketing manager Bri Driggers had a chance to talk with two industry experts about the development of AI in healthcare and what it means for our futures.

Dr. Orest Boyko has a background as a clinical radiologist and served as a residency and fellowship director for neuroradiology for several years. Almost a decade ago, Dr. Boyko decided to step back from the clinical side of things and go part-time as a radiologist for an industry AI research group in San Jose, California. The last ten or so years have opened his eyes to the possibilities of AI in healthcare, and he assures the non-believers and hesitant adopters that, in his opinion, “AI will always be a tool that is overseen by humans.”

Another expert we interviewed, John Montville, Executive Director of the Oncology Service Line at Bon Secours Mercy Health-Lourdes Hospital, also observed, “Medicine has always been an art.” Certain specialties may be more objective and less ‘artistic,’ which puts them at the top of the list of those believing AI is here to assist. However, John assures providers that even the more cut-and-dry specialties like pathology can benefit from AI without anyone worrying about becoming expendable.

Ron echoed these sentiments stating, “The only ones who are truly at risk for being replaced by AI are those who refuse to learn, utilize, and optimize this rapidly emerging tool.”

The current landscape of healthcare and AI

Ultimately, healthcare is an industry where AI lends itself far more to being “assistive,” meaning it is primarily used with human supervision.  With Dr. Orest’s radiology background, he identifies existing AI tools as “a radiology fellow in a box.” he paraphrases the statement by Dr. Nick Bryan, retired Chairman of Radiology at the University of Pennsylvania, who, during his 2022 keynote speech at the 60th American Society of Neuroradiology (ASNR) meeting, referred to the performance of his neuroradiology clinical decision tool, Galileo CDS, as a “radiology fellow in a box”. AI serves as a resource for the provider to offer better patient care via improved AUC (Area Under the Curve) performance and faster turnaround times. Additionally, mammographers who have adopted AI in their practice have recently commented on how their Mammography Quality Standards Act (MQSA) scores improved.

“Interacting with the AI tool used in my radiology practice has made me a far better radiologist,” Orest commented when asked about his opinion on the developing technology. Given his tenure in the field, he notes that many tools he now uses in his radiology practice were not taught during his residency and fellowship. Radiologists are always in a constant mode of learning and adopting innovation for their patients. For example, the tool he has been using most recently assists with the detection of lung nodules. This tool helps him to more quickly locate nodules that may have taken more time on his own to ultimately observe. Over time and with consistent use of the tool, Dr. Boyko has gathered more subtle findings that assist his pattern recognition skills for future cases.

In addition to Dr. Boyko’s insights, John pointed out that “AI is both a fellow and a mentor in some cases.” With John’s oncology background, he looks at AI through a slightly different lens.

Ron notes, “Right now, we have a radiologist shortage in rural US that is leading to multi-hour delays in some Emergency Rooms. Without a full staff of radiologists, the turnaround time for routine imaging services like CTs and MRIs is exceeding industry standards and patient expectations. If AI is adopted in cases like this with routine and normal findings, it can speed up the process to make ERs efficient while simultaneously assisting with the radiologist shortage that is projected to increase over the next decade.”

John highlights the benefits of using AI as a “second set of eyes” in breast cancer screenings, as well as further developing risk assessment tools. Could AI offer a way to assess patients and provide higher or lower probabilities of breast cancer? The short answer is yes.

Are we ready to rely on that and recommend some patients get mammograms every 10 years instead of three due to lower probability? That answer is not so clear.

While the adoption of AI in radiology has moved along a little more quickly, mammography has been slower to lean into this new technology. This is, in part, due to a lack of targeting by AI software companies. The other part of slow adoption is due to the fear of new technology, lack of adequate reimbursement, and unknown legal risks.

Healthcare is a business and one of the largest economic drivers in the US. Some providers and hospitals may be hesitant to adopt an AI tool that cuts down on revenue from procedures, office visits, and the need for more frequent routine assessments. The implications of risk assessment also involve the application of medical ethics standards and can get very complicated very quickly.

Considering AI in value-based healthcare

AI already has a history within radiology and medical imaging, acting as a “second set of eyes,” working alongside radiologists to identify areas of interest during the screening of mammograms. However, its potential extends far beyond Computer-Aided Diagnosis (CAD). Medical imaging is exceptionally well-suited for AI to play the various roles mentioned earlier:

  • Mentor: AI can aid in training new healthcare providers by offering guidance and practice in reviewing images and identifying changes through simulated cases and data.
  • Fellow: It can act as a fellow radiologist, providing valuable second opinions and clinical insights.
  • Provider: AI can handle the initial review and manage the less complex cases, allowing healthcare providers to focus on more challenging cases.

Additionally, AI can play a crucial role in “combination” medicine, where it integrates imaging with other diagnostic tests that were previously conducted separately. For example, it can add coronary artery calcium scoring to lung nodule imaging to detect clinically unsuspected cardiovascular disease. These extensions can enhance disease detection, improve patient health outcomes, and reduce healthcare costs by streamlining care and testing.

This approach extends to other imaging areas, such as interpreting Low Dose CT lung screening images and detecting pancreatic cancer at an earlier stage due to its higher specificity in identifying changes and abnormalities. When combined with genetic and family history data, AI has the potential to revolutionize early cancer detection, making treatment far more effective.

This is where the concept of value-based care comes into play. Value-based care primarily focuses on improving patient health outcomes. This supports investing in AI tools and other developing technology across all specialties to continue down the path to achieving value-based care.

So, what does the next decade of healthcare look like with AI?

John notes that “AI can create far better clinical pathways and extremely precise treatment recommendations, and that’s something worth investing in. It will create revolutionary change in the creation of true precision medicine, allowing the clinician to better organize all aspects of the patient’s health state and all data on the cancer type and makeup to create a treatment plan that is built for that patient and their cancer care.”

AI and the digitization of pathology departments have a similar path as we’ve seen in AI and radiology.  Both have strong focuses on the concepts of “Assistive” and “Augmentative” and are well suited to more “black and white” or binary evaluations. In pathology, the stain either turns red or it doesn’t, and counting cells per specimen is as objective as a result can be. In imaging, the spot in the lung (for example) has either increased in size, decreased in size, or remains the same. These are all observations that have little room for subjectivity.  


The integration of Artificial Intelligence (AI) into healthcare is currently a topic of intense interest and debate. While concerns about AI replacing healthcare professionals exist, experts in the field emphasize that AI is designed to complement human expertise rather than replace it. The current landscape in healthcare sees AI serving primarily as an “assistive” tool, enhancing patient care and speeding up diagnostic processes.

AI has potential applications in value-based care via developing precise treatment recommendations, creating more effective clinical pathways, and enabling the customization of treatment plans for individual patients. Overall, the next decade in healthcare with AI promises revolutionary changes that align with the value-based care approach, enhancing patient health outcomes and the quality of care across various medical specialties.