Key Takeaway:
- AI-assisted test by Mayo Clinic detects pancreatic cancer up to three years earlier, improving chances for curative treatment.
- A study of nearly 2,000 CT scans shows AI identifies early cancers at almost double the detection rate of specialists.
- Breakthrough supports a shift toward preventive, early-detection medicine using artificial intelligence
The Mayo Clinic announced Wednesday an AI-assisted test that detects pancreatic cancer up to three years before diagnosis by analyzing CT scans, potentially improving survival by enabling earlier, curative treatment.
AI Model Identifies Cancer Before Tumors Form
Researchers said the artificial intelligence model detects subtle biological signals of pancreatic cancer even when scans appear normal to specialists.
The new testing system analyzes routine CT scans and identifies patterns linked to cancer development long before tumors become visible. Early detection is critical because pancreatic cancer often shows symptoms only in advanced stages, limiting treatment options.
“The greatest barrier to saving lives from pancreatic cancer has been our inability to see the disease when it is still curable,” said Dr. Ajit Goenka, a Mayo Clinic radiologist and senior author of the study. “This AI can now identify the signature of cancer from a normal-appearing pancreas.”
Experts said the technology could give patients years of additional time to pursue surgery or other treatments aimed at curing the disease rather than managing late-stage illness.
Study Reviews Nearly 2,000 CT Scans
The findings, published in the medical journal Gut, are based on an analysis of approximately 2,000 CT scans, including images from patients later diagnosed with pancreatic cancer.
Researchers reported that the AI-assisted model detected prediagnostic cancers at nearly double the rate of physicians reviewing scans without AI support. The technology also proved especially effective when analyzing scans taken more than two years before an official diagnosis.
According to the study, AI-reviewed scans were three times more effective at identifying early cancers compared with traditional review methods. Scientists said the results demonstrate the potential for AI to transform cancer screening by uncovering disease earlier than previously possible.
Independent medical experts say earlier diagnosis could significantly improve survival rates, as pancreatic cancer remains one of the deadliest cancers due to delayed detection.
Research Supports Preventive Medicine Initiative
The project was conducted as part of the Mayo Clinic’s Precure initiative, a program designed to predict and prevent disease by studying biological changes before symptoms emerge.
Researchers said the initiative focuses on shifting medicine from reactive treatment to proactive prevention. By identifying early warning signs hidden in medical data, physicians hope to intervene before the disease progresses.
Mayo Clinic officials said further clinical validation and regulatory review are expected before the AI model becomes widely available in hospitals. Additional studies will examine how the technology performs across broader patient populations and health systems.
Health care leaders say the development reflects a growing role for artificial intelligence in medical imaging and diagnostics, where computers assist physicians rather than replace them.
“This is about giving clinicians better tools,” Goenka said. “Earlier knowledge means better decisions and, ultimately, more lives saved.”
