An innovative new test, powered by artificial intelligence, can detect pancreatic cancer up to three years before diagnosis, offering hope for increased survival in one of the deadliest forms of the disease.
Researchers from the Mayo Clinic have developed a model called REDMOD, which identifies very small changes in tissue that are usually not discernible by traditional imaging methods or by radiologists themselves.
Pancreatic cancer is considered particularly dangerous because of its “silent” nature. In its early stages, symptoms are vague – such as mild back pain, fatigue or digestive problems – and are often overlooked. As a result, about 80 percent of cases are only diagnosed after the disease has spread, when surgical intervention is no longer possible.
The study, published in the journal Gut, showed that REDMOD was able to detect signs of cancer an average of 475 days before diagnosis. The system was more accurate than radiologists, identifying 73 percent of cases compared to 39 percent that specialists achieved.
According to lead author Ajit Goenka, the biggest challenge has always been identifying the disease at stages when it is still treatable. He points out that this technology can detect signs of cancer even when the pancreas appears normal in appearance.
Experts believe this approach could change the way pancreatic cancer is diagnosed, moving from late detection to early preventive intervention. However, further studies are needed to confirm its effectiveness in widespread clinical use.
Currently, only about 12 percent of patients survive five years after diagnosis, making any progress in early detection an important step in the fight against this disease. /GazetaExpress/
t7 live
