Arun Hariharan, MD, began using an artificial intelligence tool that streamlines spinal imaging for his pediatric patients.
He leverages BoneMRI, which creates a CT-like image from an MRI scan. Dr. Hariharan, of Paley Orthopedic and Spine Institute in West Palm Beach, Fla., spoke with Becker’s about the advantages of BoneMRI and how the AI tool is improving the patient experience.
Question: Bone MRI — this AI tool that you’ve been using — can you talk more about that, how long you’ve been using it, and some of the outcomes that you’ve gotten from it?
Dr. Arun Hariharan: We just got it a couple months ago, and it has been super helpful. Every patient that we want to see the bony and soft tissue details on we get one study now, instead of an MRI and a CT scan. Especially in children, that has been very beneficial, both in terms of cost for the patient and the families. Instead of coming twice to get two different studies they get one study, and from a radiation standpoint, it reduces exposure.
The cool part of it is that the AI makes the MRI look like a CT scan. An MRI is very good for the disc and spinal cord and nerve roots and muscle. A CT scan is very good for the bone, but an MRI is very underwhelming for bone. You can’t see the bony details very well, and on a CT scan, you can’t see the soft tissue details very well.
Q: What are the cost savings that are associated with this technology, just like any numbers you can share on that.
AH: I don’t know the actual numbers of it, but from a logistical standpoint, you don’t need to have the patient scheduled for two different studies. You don’t need to get insurance authorization for two different studies. From a time standpoint, the parents don’t have to bring their children for two different studies. So besides the monetary parts of that, there’s just lots of overall quality of life and ease of use benefits.
Q: How else do you see this type of technology evolving in the spine space?
AH: As with anything in any space, AI is gonna have a big role. So far it’s very good at converting the spine images into CT-like images, and I think it’ll expand beyond the spine. It already has in the pelvis and some extremities. I think in the future, we likely will not ever have a need for a CT scan, which has been around for a long time.
Q: How are you thinking about safeguarding AI and keeping precision?
AH: That’s a great point, I think in most of these cases we’re trying to prove a diagnosis that is often better proven with CT scan. It’s beneficial if you’re planning for surgery. I personally haven’t used it in a navigation setting, but it exists where you can use this system to navigate this group and be pretty effective. But as with any AI, clinical judgment and understanding of limitations is important.
Q: How else have you been using AI in your practice this year?
AH: I’ve been using it in one form or another for several years, so this is the latest part of it. But there’s, for example, patient specific rods, which are sort of AI driven. Human engineers design the rods, but the parameters and the technology behind it is AI driven. We also use AI for It’s just all over the place.
