3/3/26

Using AI to Find Buried Signals

So I think this all comes down to really understanding patient phenotypes. And I think there's been incredible advancements in the field of oncology and being able to do really specific, you know, genomic signatures of tumors of different biomarkers and get hyper, hyper precise in terms of, you know, the type of cancer that a patient might have and then develop really, really innovative therapies that can target those specific, you know, mutations.


A lot of times as people think about like cardio metabolic disease, people think of these as like these huge chronic buckets, right? There's heart failure. But the reality is, know, heart failure can manifest in so many different ways. You know, within heart failure, you obviously have, you know, preserved ejection fraction, reduced ejection fraction, etc. But even beneath that, there's so many different comorbidities or biomarkers that these patients have that create a very, very heterogeneous and complex disease. And so as we think about the opportunity to really bring that type of precision approach to cardiology, I think it's massive.And really what we're starting to see is maybe less on the kind genomic biomarker side as you see in oncology, but more on kind of what are the different signatures that show up from imaging, right? 


The echocardiogram, the ECG, chest CT, CCTAs. There's a ton of really rich information that is captured in these images. And, you know, we're really only scratching the surface of the insights that we can glean from them - there's a huge opportunity to bring that precision medicine approach that's been so revolutionary in oncology to larger chronic conditions like cardio metabolic disease, respiratory disease, renal disease and really start to unearth insights from imaging things like ECGs, echocardiograms, chest CT, CCTAs that can start to showcase different signatures and biomarkers that are relevant for these diseases. 


Ultimately, we see a vision where within cardiovascular disease, we're able to substratify phenotype patients at a much, much more granular level to be able to really understand where are they at within their disease progression? How are different comorbidities affecting potential outcomes? And how do we treat that patient with a much more precise approach versus thinking of it as just a large cardiovascular disease patient?


Previous

Bringing Precision Approaches to Cardio Trials

Next

Bringing Precision Approaches to Cardio