Yair Lotan, MD, on the integration of AI into urologic care

In this interview, Yair Lotan, MD, recaps a session from the 2024 Bladder Cancer Advocacy Network (BCAN) Think Tank titled, “Harnessing the Power of Artificial Intelligence in Advancing Bladder Cancer Research and Clinical Care.” Lotan is the chief of urologic oncology and a professor of urology at UT Southwestern Medical Center in Dallas, Texas.

Video Transcript:

There's definitely a lot of hype around the concept of artificial intelligence, and I think we're in the early phases of trying to figure out where it makes sense to adopt the technology and how to incorporate into clinical care. The whole concept of artificial intelligence obviously involves a fairly wide spectrum, from having a computer teach itself a language and try to solve complex mathematical algorithms to composing poems. But when you try to look at clinical care, you really need to identify areas with a focus of how it's going to actually improve decision-making or help clinicians think about certain problems, or possibly even help fine tune predictions about patient outcomes. You can imagine that certain areas are a little bit more ripe for AI incorporation, so think about pathology or radiology, where you have multiple images with recognizable patterns, where you're trying to ask it, do you see any abnormal findings? If you can identify, for a program, what the abnormalities would look like typically, then it can help highlight those areas for radiologists or for a pathologist, for example, so that it can hopefully augment what they do.

The concern, of course, is should it just replace what they do? But I think we're far away from that right now. Though, there are already some artificial intelligence programs and companies that are involved in improving diagnostics, for example, or prediction in prostate cancer. So, I think this is going to be a field which is going to grow. And I think the main focus for clinicians is one to make sure it's robust––in terms of design of the programs, in terms of utility––and to make sure that not only is it helpful, either saving time, but also improving accuracy, improving patient outcomes. So, a lot of the session focused a little bit about the mechanics of how artificial intelligence works, and then also looked at some specific examples where there are already some evidence that artificial intelligence will improve or could potentially improve patient care.

This transcription has been edited for clarity.