ISC 2026 | The current use and future opportunities for AI and novel trial designs in acute stroke care

Well, in terms of trial designs, we’ll take first, we already use a lot of novel approaches that we didn’t use 10 years ago. A lot of what we use is something called adaptive designs, which you, you set up rules when you start a trial, and you say, if this happens, we’re gonna do that because it’s not as kosher to get halfway in the trial and say, hey, we ought to change this if we see this, that’s not, because that leads to maybe overestimates of what you’re trying to find with your treatment. So we adapt, and this adaptation really leads to advances more quickly. So I think that’s one major thing. The second thing is we’re using what’s called a step platform, which is you have a platform built for, we have now for acute stroke, reperfusion, and we can add things in and take things out on an ongoing basis, so it can be more efficient, so we don’t have this lag time. We can keep adding and subtracting things as they get tested. And so there’s some unique ways we do that, but we’ve had platform trials in cancer in other areas, but we’ve not had one in stroke until this last year. There’s also another platform trial that’s based globally that we work with as well. So I think those are two very exciting areas that trial design has evolved and they’ll be different ones. I think we’ll also be more pragmatic, trying to test things in the flow of the daily treatment of patients, whether it’s in the clinic or the hospital, and these pragmatic trials can test things that we’re doing, but we’re not sure which of two things is better, and that can also lead to major advances and incremental advances in what we do.

The use of AI that is also got a much longer, I would say, time frame that it’s gonna happen. But I do think we’re starting to, we already use it right now for imaging to help decide if there’s a certain eligibility of a clot seen in an artery or a certain size of brain hemorrhage where we can use imaging now and AI to help us do that. But I think what we’re talking about in the future is somebody comes in. You get their information. And an AI partner can get all the details about that patient. We know then what the situation we’re trying to treat. The AI pulls all the available information from clinical trials that are relevant for that person and then says, here are the options, here’s the data, and the physician partnering with AI can say, this is what’s best for you, because when we do trials, we look at the overall effect in the whole population, but some people in the trial may do tremendously well, some may not get much response, and some may be harmed. So, if we had a way of knowing who is going to be likely harmed or who is gonna be helped a lot, we could then select that particular treatment for that individual patient. More like the precision medicine we’ve been talking about, but we haven’t quite gotten there yet.

This transcript is AI-generated. While we strive for accuracy, please verify this copy with the video.