VIDEO: AI, big data may transform CLL treatment, prognosis

Editor’s note: This is an automatically generated transcript. Please notify editor@healio.com if there are concerns regarding accuracy of the transcription.

I think this is one of the most exciting opportunities for artificial intelligence. I mean, I get criticized a little bit that I'm gonna develop a Skynet for healthcare, right, with artificial intelligence. Personally, I think in medicine, the opportunities for artificial intelligence are monumental. So I just mentioned that big C word, you know, cure, for some of these chronic conditions. I think the way we're gonna get there, frankly, is using big data and basically, looking at some of these mutations that have been really ignored, frankly. We have a lot of things called VUSs on our NGS panels, right, variants of uncertain significance.

We can start making those significant when we combine large database studies with our molecular studies. And I've developed a protocol here at Roswell Park, and I'm sure others are working on this actively elsewhere, where we're actually working to go back in time, look at how people have done, looking at their molecular features, looking at some of the clinical features, and you really create this comprehensive model where you can use machine learning algorithms to train using all these small data elements that we oftentimes will overlook. And you can actually develop novel prognostic and predictive algorithms, which can help you almost predict probably to the month of when someone can have recurrence over time.

We could probably develop rational combination treatments, which use even, like, new antigen-based approaches or other immunologic parameters, like cytokine profiles even, that can help predict resistance to therapy. And so I think this is gonna be the future. This is how we're gonna get to cures, is using complicated models that are trained with real-world data from the past, and then we validate them prospectively going into the future. And in my opinion, this is gonna lead to the cures for all cancers, frankly, over time, provided enough data, clinical support, you know, research support, and using novel therapies in rational combinations.

So actually, frankly, I believe that the cure for all cancers will lie in these kinds of approaches, given enough time and enough research and money and work, hard work, and mutual trust between doctors and patients, right? Data is not to be taken lightly. So data security is in incredibly important in this, especially genetic information. So it can be incredibly damaging. So building secure networks and best practice and stewardship is critical to this to develop that trust and to allow these things to occur and grow over time.