What is the role of AI in the cath lab of tomorrow?

So in my opinion, AI these days is everywhere.

Uh, whether you go from, uh, industry into manufacturing, engineering, or even the way we process foods, it's everywhere.

That said, to me, in the cath lab and also in interventional cardiology, uh, by itself I think we are already implementing AI, and I would even take that one step back from the cath lab to even prevention.

If you look these days at how we screen patients for coronary artery disease using, for instance, CTA, in which we now have a multiplicity of, of, of vendors offering really cool software solutions to get automated AI-based block quantification, uh, physiology, even without minim- uh, minimal, uh, operator input.

That's, that's a great example on how AI can be used to, to guide decision-making, to guide risk stratification, and to guide even, uh, complex PCI.

Within the cath lab, I think we are, uh, on the verge of integrating a lot of AI-based technology.

Think about imaging processes, for instance, upgrading your, your angiogram by using AI-specific filters.

Uh, you could think about automated plaque quant- uh, quantification in patient with IVUS or OCT analysis, but also, for instance, completely automated angiographically based, uh, physiology assessments.

So there are a lot of tools already that are, uh, currently being used, uh, or can be used in, in patients with suspected or even, um, more precisely defined coronary artery disease.

Okay.

Yeah, fascinating.

And, uh, how do you imagine, uh, a term, uh, in the future, uh, the full integration of AI in our cath labs?

Yeah, so that's actually perhaps the even more interesting question because every- everybody obviously recognizes there's a lot of tools currently being introduced that could help us.

But I think to me, the challenge is that these tools are there, and they are, they are technically validated, but I think there are still a gap in between this technical validation and how these tools finally will be used in order to really improve patient outcome because that translation from, from bench testing and industry to, to a large-scale outcome trials currently, I think is, is missing.

Uh, and that is something that, uh, that is, that is I think hampered by a lot of challenges from ethical concerns, the ability to use large-scale clinical data sets, and also to, um, yeah, actually to, to allow guideline interpretation of, of, of embedment of these, these novel tools.

And, uh, how could we prepare to, for this time of, uh, full AI integration?

Yeah, that's a, that's a, that's a difficult one.

I think, uh, obviously a very close collaboration between industry partners, engineering, uh, AI specialists, and the clinical field, so the, the interventional cardiology, uh, community I think is, is needed.

Um, I think the best developments and the best, best solutions for, for guiding treatment are not just based on a, on a technical idea.

They're usually based on a, on a clinical need.

There needs to be a, a need introduced by the, by the clinical community, um, in terms of, of, of what is missing.

For instance, the, the amount of, uh, inflammation or lipid in a plaque versus outcome.

And then I think the, the, the collaboration between the industry partners and the clinicians is critical in order to come to the best solution.

First, to quantify it, and then second, um, even beyond the individual, uh, interventionalists, the, the societies, the communities are, are, are also needed in order to allow the use of mass clinical data in order to set up large-scale registries, clinical programs that are feasible, uh, in a, in a way to, to validate these, these tools.

So I think that is the, that is the challenge, uh, which imposes obviously, uh, technical concerns, uh, ethical concerns.