For healthcare stakeholders and regular observers, it’s understandable if the hype around AI might seem a bit much, at times. After all, it’s not as if developers of new technologies have a habit of underselling their product’s possibilities.
With AI, however, it may be that the best antidote for the possibility of developing a jaundiced eye about AI may be to take a look at some of the, well, really cool applications to which the technology is already being put to use.
Writing recently at HITConsultant, for example, Ron Soferman, founder and CEO of RSIP Vision, a computer vision and deep learning consultant, provides just such an antidote with a look at three current AI use cases.
For starters, Soferman describes how AI is being tapped to help doctors planning and carrying out cardiac interventions.
“(U)sing arrhythmia as an example,” he explains, “you can create a map of the heart to pinpoint the exact problem causing an irregular heartbeat. Mapping provides the exact anatomical structure of the arteries. This is a major assist in planning interventions with a catheter. Mapping can determine the exact kind of catheter to be used and the exact behavior of the arteries at the specific point where you have to do the intervention. Mapping sometimes can be used during an operation itself. With images from the fluoroscopy on hand, AI provides analysis of those images in order to get timely and precise information about the location and the structure of the arteries.”
As if to demonstrate the versatility of AI, Soferman turns next to a look at how AI is being used in spinal surgery. The tricky part, he says, is when screws need to be put into the vertebrae, a move made more difficult when the actual insertion is unseen due to it being a percutaneous procedure.
“Utilizing pre-op scanning, along with information provided by the x-ray, artificial intelligence assists in the operating room by detailing exactly where the vertebra line up. It happens through algorithms combining those two sources of information which allow the surgeon to accurately navigate to the exact point of insertion.”
Finally, switching tracks altogether, Soferman points to how, in pharmaceutical development, AI is being used in assessing the influence of new drugs.
“A lot of effort is invested by collecting CTs that are done for the patient during the use of a new drug,” he says. “AI algorithms take all of that information to produce an automatic scoring (which is called RECIST) that analyze and measures the impact of the drug (an example being whether a lesion has disappeared or shrunk or whether it stayed the same). Through AI, results of these scans can be determined within a few hours, as opposed to several months as before, providing better and more immediate decision making.”
If there is a downside to current uses of AI, Soferman suggests it’s due to an ongoing “shortage of trained engineers that can develop new algorithms to solve ever more difficult situations.”
That said, Soferman believes the last five years have brought a new understanding of what AI can do right now, which will spur even more integration of newer and better AI technology into the healthcare system moving forward.