Mayo Clinic's New AI Tool Could Transform Pancreatic Cancer Diagnosis

Reported by MindBodyGreen.

Pancreatic cancer doesn't kill because it's untreatable — it kills because it's invisible until it's too late. More than 85% of cases are diagnosed after the disease has already spread beyond surgical reach. By the time a tumor shows up on a scan, most options for a cure are off the table. There's no standard early-detection protocol for the majority of cases, so most people don't find out until symptoms arrive — and at that point, the window is effectively closed.

What AI Can See That Radiologists Can't

A research team at Mayo Clinic built a tool specifically designed to close that window. REDMOD — short for Radiomics-based Early Detection MODel — doesn't hunt for a tumor. Instead, it reads the tissue itself, identifying microscopic textural disruptions in the pancreas called radiomic signatures that suggest cancer is developing, even when everything looks completely normal to the human eye. According to MindBodyGreen, 90% of REDMOD's predictive features come from specialized filtered images that capture structural changes no radiologist could detect on a standard scan.

The results, published in Gut, are striking. REDMOD correctly identified 73% of pre-diagnostic pancreatic cancers — nearly double the 38.9% detection rate of radiologists. More critically, it flagged those cancers a median of 475 days before clinical diagnosis — that's roughly 16 months of lead time. For cancers more than two years out from diagnosis, the AI hit 68% sensitivity versus just 23% for human reviewers. Its consistency across repeat scans topped 90%, and it accurately identified healthy pancreases in both a multi-institutional dataset (81.3%) and a public NIH dataset (87.5%).

The group with the most to gain right now: people who develop diabetes after age 60. New-onset diabetes carries nearly a 20-fold higher risk of pancreatic cancer compared to the general population — a risk window that current imaging largely fails to act on. REDMOD could become a critical layer in high-risk screening protocols, catching cellular changes before a visible tumor ever forms. Other elevated-risk profiles include those with a family history of the disease or a history of chronic pancreatitis.

REDMOD is not available outside research settings yet. It's moving into real-world clinical trials under the study name AI-PACED (Artificial Intelligence for Pancreatic Cancer Early Detection), which will test how the model performs in practice. Researchers estimate that even modest improvements in early detection could more than double survival rates — a number that puts the stakes of this technology in sharp relief. If you carry known risk factors, the most useful thing you can do right now is talk to your doctor about what monitoring makes sense for you.

The bottom line: AI may finally give us a way to catch one of cancer's most lethal forms before it becomes a death sentence.

Read the original at MindBodyGreen.