Scott Ceresnak, MD, cardiac electrophysiologist, director, of the Pediatric Electrophysiology Program, Stanford Medicine Childrens Health, Palo Alto, California, shared his excitement regarding the potential of artificial intelligence (AI) technology in the pediatric cardiology space, including wearable technology, in this video interview with Contemporary Pediatrics.

Ceresnak noted mass research is needed to figure out how AI can further be used in health care, but explained trials are underway in various health care spaces.

Transcript (edited for clarity):

Contemporary Pediatrics:

Dr. Ceresnak, when it comes to AI, it's dominating the headlines in the medical field for a varying number of reasons, but we have seen in some indications and populations, that technology trickle into the pediatric space. From a cardiology perspective, what are some of those findings and what are you starting to see be rolled out, or tested if nothing else?

Scott Ceresnak, MD:

You know, AI is such a hot topic now. And, really, there's so much promise with the technology going forward. I would say a couple of things with AI, first of all, as you mentioned, so much of the AI data that's being generated, and algorithms that are being generated are adult centric, right?

So I think we have to be a bit careful. And just because these algorithms work well, on adults, we don't know if that's going to trickle down if those specific algorithms are going to work in children. So I think the first step is, for most AI algorithms, you need big data, right? So you need a big data sets to understand how can you distinguish normal or abnormal. Then you need a good big group to validate that algorithm. Does it work? So, we need big datasets in children that can really, A.) Provide good machine learning algorithms, and then B.) validate them and see do they work or not.

I think that's a big step for us is really getting a lot of big data to generate these algorithms and generate specific pediatric level algorithms. So I think the first steps that are coming is do these algorithms work in children? Do the adult algorithms working children? And then can we develop our own pediatric specific elements? And I think there's a lot of work in CG space, electrocardiograms, and that there's a lot of work that a bunch of groups are doing here at Stanford and a bunch of other groups across the country, great centers are doing a lot of machine learning work on ECGs.

As I mentioned before, echocardiography, looking sonograms of the heart, one of my colleagues at Stanford Charitha Reddy is doing great work looking at, can we create artificial intelligence-based readings of these echocardiograms? Can we distinguish normal from abnormal and take away the human part of the reading. And really, I think amazing work is being done. I feel like we're sort of at the tip of the iceberg and what we can do and what we're starting to see in the pediatric space and I'm super excited about what's to come, really on the electrophysiology side and ECG algorithms and other arrhythmia detection algorithms. But then also, looking at in hospital infection of arrhythmias and early detection of potential events in the hospital, and imaging wise. I think, on the imaging side, this is going to revolutionize the way we sort of see echocardiograms and other imaging modalities in children.

I'm excited about this space, especially wearables, and I think they carry great promise and I think they carry a lot of hope for arrhythmia detection in children. I would caution people too to say that, as I mentioned before, all these algorithms are built for adults.

I think we have to be careful how we interpret the data and the combination of using the watch and using the data that we can get from these smartwatches, but also using the clinical judgment of important people who are experts in the field, is going to be really important to strike that balance between not raising anxiety too much in terms of families, but also, truly picking up the arrhythmias that we can treat and we can help kids going forward.

Follow this link:

With the correct research, AI in pediatric cardiology has great potential - Contemporary Pediatrics