Editor's Notes: What the Trajectory of AI in Radiology Says About the Unexpected in Healthcare

As I reported from McCormick Place in Chicago last week while covering RSNA24, “Though progress is turning out to be very complicated in multiple ways, the adoption of artificial intelligence (AI) is definitely rolling forward across the medical specialty of radiology, as speakers at RSNA24, the annual conference of the Radiological Society of North America, documented in session after session on December 2 and 3.”

I noted that “A number of sessions at this year’s RSNA Conference, being held at Chicago’s vast McCormick Place Convention Center, focused on practical innovations taking place in patient care organizations nationwide right now. There seem to be nearly an unlimited number of possibilities right now, speakers emphasized; the opportunity lies in focusing on specific areas of potential and moving forward thoughtfully and strategically.”

Indeed, I noted last Monday that “In a highly stimulating lecture, Eric Topol, M.D., a bestselling author and a practicing cardiologist at the Scripps Clinic in San Diego and editor-in-chief of Medscape, told a standing-room-only audience at the plenary session on Dec. 2 at RSNA24—this year’s conference of the Radiological Society of North America—that artificial intelligence will transform the practice of medicine in the coming years. Speaking to a standing-room-only audience at the Arie Crown Theater in Chicago’s vast McCormick Place Convention Center, Dr. Topol, author of the 2019 bestseller Deep Medicine, walked his audience of radiologists and others involved in radiology, through the evolution to date of artificial intelligence, and then predicted based on progress so far, what will happen next.”

Dr. Topol spoke boldly and confidently, predicting that medicine is on the cusp of being able to make use of two types of multimodal AI—one based on text, speech, and images, and the other based on human data. “Where can multimodal AI take us?” he asked his audience last Monday. “You can get into a much different level of precision and accuracy medicine going forward,” he predicted. “For example, hospital-at-home can be contemplated more in the future,” as the analytics needed to support such leading-edge care delivery forms will more and more be available.

And I wrote that “[W]hat seemed clear this year is the nearly unlimited range of possibilities, clinical, clinical-operational, and operational, across the specialty. Broadly speaking, radiologist leaders are focusing on a few overarching areas: AI to support initial diagnostics; AI for clinical decision support around type of diagnostic test to order; AI to support intelligent scheduling and protocoling; the use of large language models to support patient record and history summarization; and the use of LLMs to facilitate the translation of radiology reports and information into patient-friendly language and framing.”

I must say that the atmosphere at RNSA24 was vastly different from what it had been even five years ago at RSNA19, when there was still lingering apprehension among many radiologists that the adoption of AI would eliminate some of their jobs. In fact, no jobs have been eliminated; instead, radiologists have come to understand that AI adoption will be of tremendous help to them, as they attempt to keep up with a snowballing level of demand for diagnostic imaging services, given the aging of the population and an explosion in chronic illnesses. Indeed, as one expert said recently, “It’s only the radiologists who refuse to use AI who might lose their jobs”—meaning that AI is rapidly becoming a must-have technology, for its decision support, study prioritization, diagnostic support, and other capabilities.

In his presentation last Monday, Topol referred to “Machine Eyes”—the collection of data that, when analyzed and poured into clinical decision support, can improve diagnostics. And he noted that the foundational work over the past numerous years in developing algorithms and working with large language models, has set the stage for massive change. For example, he told his plenary audience, the data gathered from enormous amounts of data and images, is already leading to better diagnoses, as in the case of gastroenterology, where gastroenterologists are already using AI-facilitated endoscopy to achieve detect more polyps than they could previously. And data is being gathered even from such diagnostic images as x-ray, creating massive lakes of data that are being used to support physician diagnosis processes.

And beyond the purely clinical, Arun Krishnaraj, M.D., M.P.H., a professor of radiology and medical imaging at the University of Virginia, noted on Tuesday that he and his colleagues at UVA are already leveraging large language models to help them recast their radiology reports in laypeople’s language, enhancing relationships with patients and strengthening patients’ engagement with their own healthcare delivery.

So what all the speakers in all the sessions focused on AI at RSNA agreed on, was the fact that we are still very, very early in this journey around AI in radiology and in other specialties, including cardiology, gastroenterology, endocrinology, dermatology, and nephrology, among others. So it is very difficult to predict specific outcomes for all of this early work; already, in fact, there have been significant surprises.

And it’s heartening that the physician leaders involved in all this work, in radiology and in all the other specialties in which work is proceeding apace, are moving forward thoughtfully and strategically, looking to create early wins that will instill confidence among their fellow clinicians and among patients, while also investing time and effort to create good governance systems. The fear-based image that some had a few years ago, of crazed “mad-scientist” types going berserk on the technology with no controls, is not playing out at all; instead, leaders are moving ahead intelligently, while focusing on strong early wins. It’s as it should be. And yes, we should simply expect surprises along the way. But perhaps the biggest has been the extent to which the adoption of AI is already helping radiologists become more efficient, effective, and even accurate. And that is a hopeful sign indeed.