Q: Speaking of how technology can empower healthcare professionals, how do you see the impact of generative AI on radiology and on healthcare in general?
A: I think its impact is going to be huge; there is going to be a pre- and post-generative-AI world in healthcare. One of the biggest opportunities I see is to bring various sources of patient information together in a meaningful way, to paint a more complete picture of that patient – which is something we have always struggled with in radiology because of how fragmented healthcare is.
Today, we tend to look at diagnostic imaging in a silo. The vast majority of imaging systems sits in its own realm, disconnected from the EMR, lab results, and other patient data. As a radiologist I can tell you what I see in a MR or CT scan, but the more I know about the patient, the better I’ll understand what I’m looking at, and the more value I will be able to provide. That’s where generative AI can support me, by pulling different information sources together, and then generating a list of potential diagnoses and corresponding probabilities for a particular imaging study.
UItimately this will pave the way towards what I like to call ‘ambient intelligence’, where I sit down at my radiology workstation and have access to every bit of information I need at my fingertips. Not only that, but a virtual assistant will also provide me with relevant insights and recommendations. It may say to me: “Dr. Pitt, I have taken a preliminary look at this patient’s images, I’ve also consumed information from the EMR, lab results and other sources, and I’ve found a similar patient who is also 73 years old, with diabetes and heart disease. Based on my analysis I expect to see these and these comorbidities, and this is something I suggest you pay close attention to.”
So you’re going to see increasingly intelligent support systems, but the radiologist will remain in control and accountable, much like the pilot of a plane.