Researchers have developed an artificial intelligence (AI) model that could help identify people at increased risk of pancreatic cancer up to five years before diagnosis, potentially creating new opportunities for earlier detection.
The model was developed using electronic health records and routine blood test results from almost 40,000 people. By analysing patterns in patients’ medical histories over time, the AI identified signals associated with a higher risk of developing pancreatic cancer. These included changes in blood test results, diabetes, pancreatitis and other health conditions.
Pancreatic cancer can be difficult to detect in its early stages, and many people are diagnosed only after the disease has spread. As routine screening for pancreatic cancer is not currently practical for the general population, identifying people who may be at higher risk could help doctors determine who might benefit from further investigation.
The initial findings suggest that the model could distinguish between people at higher and lower risk of developing the disease.
Researchers are now planning further testing to assess whether the model could be used in clinical practice. Important questions remain about what follow up tests should be offered to people identified as higher risk, and how to balance the potential benefits of earlier detection against the risks of unnecessary testing and false-positive results.
If further studies confirm its effectiveness, this approach could eventually help doctors identify people who may benefit from closer monitoring or additional tests before pancreatic cancer becomes clinically apparent.
Read more: New AI model may identify pancreatic cancer risk up to 5 years early