Mayo Clinic Develops AI To Detect Pancreatic Cancer Up to Three Years Before Diagnosis

With the pancreas buried deep in the abdomen, routine screenings are practically nonexistent, and symptoms like sudden weight loss or stomach pain usually appear only after the disease has spread. But a groundbreaking artificial intelligence model developed at the Mayo Clinic in Rochester, Minnesota, could fundamentally shift how and when the disease is caught.This new AI tool can detect subtle abnormalities on routine abdominal CT scans up to three years before a clinical diagnosis is made, according to findings published this week in the journal “Gut,” “The greatest barrier to saving lives from pancreatic cancer has been our inability to see the disease when it is still curable,” said the study’s senior author Dr. Ajit Goenka, a Mayo Clinic radiologist and nuclear medicine specialist. “This AI can now identify the signature of cancer from a normal-appearing pancreas, and it can do so reliably over time and across diverse clinical settings.”With a five-year survival rate lingering below 15 percent, pancreatic cancer is projected to become the second-leading cause of cancer-related death in the U.S. by 2030. More than 85 percent of patients currently receive their diagnosis after the disease has reached an advanced stage.To train the system, known as the Radiomics-based Early Detection Model, or REDMOD, researchers fed it nearly 2,000 CT scans from patients who were originally screened for other medical conditions but were later diagnosed with pancreatic cancer. These scans had previously been interpreted as normal by human experts.The AI model successfully identified 73 percent of those prediagnostic cancers at a median of about 16 months before clinical diagnosis. This detection rate is nearly double that of human specialists reviewing the same scans. For scans obtained more than two years prior to diagnosis, the AI identified nearly three times as many early cancers.Instead of looking for a visible mass, REDMOD measures hundreds of quantitative imaging features, analyzing tissue texture and structure to capture faint biological changes.“We knew, based on the biology of the disease, that this is not something which is coming all of a sudden in three months. … We knew that the signal was there. We just needed to find a way to be able to detect it,” Dr. Goenka said.The tool holds incredible promise for patients with existing risk factors, such as a family history of the disease or new-onset diabetes. “Unfortunately, if they have symptoms and if it’s truly pancreas cancer, you don’t need AI for that,” Dr. Goenka added.The Mayo Clinic model is currently advancing into a prospective clinical trial called AI-PACED, which will monitor patients for three to five years. “In a disease where we have been just wandering in darkness for decades, this is a milestone that shows us the finish line, but we still have to get to the finish line,” Dr. Goenka said.

Mayo Clinic Develops AI To Detect Pancreatic Cancer Up to Three Years Before Diagnosis

More from the New York Sun

Collectivists Turn May Day Into a High Holiday

U.S. Floats Netanyahu–Aoun Meeting as Path to IDF Exit From Lebanon

Louisiana Lawmakers Advance Bill Protecting College Faculty Members’ Ability To Engage in the ‘Rigorous Exchange of Ideas’

More from the New York Sun

Collectivists Turn May Day Into a High Holiday

U.S. Floats Netanyahu–Aoun Meeting as Path to IDF Exit From Lebanon

Louisiana Lawmakers Advance Bill Protecting College Faculty Members’ Ability To Engage in the ‘Rigorous Exchange of Ideas’

Assassination Attempt Exposes Danger of Liberal Bias — and Outright Trump-Hatred — at America’s Universities

The Marriage Gap Is the Biggest Cause of Inequality in America, but Gets the Least Attention

Kentucky Derby Arrives With History, Heartbreak, and Change in the Air

Actress Takes a Beating From Harvey Weinstein’s Lawyers Over Secret ‘Love Note’: It’s Latest Example of Why This Rape Case Is So Hard To Prove

A Tale of Moral Courage in World War II, Chronicled Through Cartoons