Mayo Clinic just published a validation study on a model called REDMOD, the Radiomics-based Early Detection Model, that spots pancreatic cancer on routine abdominal CT scans up to three years before clinical diagnosis. On the validation set, the AI caught 73% of cancer cases. Specialist radiologists looking at the same scans caught 39%. The median lead time was about 16 months, with some catches happening 475 days before symptoms.

Pancreatic cancer is one of the worst diagnoses you can get, mostly because by the time it’s found, it’s usually too late. Five-year survival is around 12%. The whole game is early detection. A model that almost doubles what trained radiologists can see, on scans that were already taken for other reasons, is the kind of result that actually moves the needle.

We all know AI is going to accelerate healthcare. The hard part is the gap between a great study and something that’s actually running in your local hospital. News like this makes it feel like that gap is closing. The next step is a prospective trial called AI-PACED, which will test how clinicians integrate this into real care for high-risk patients, like those with new-onset diabetes.

My hope is that the path from validated paper to widely deployed tool gets shorter. Pancreatic cancer is one of those diseases where every month of earlier detection translates directly into people who get to live. The technology now exists to do this. The question is how fast we can get it to the people who need it.

Link to the article