American researchers have presented an artificial intelligence model capable of detecting early signs of pancreatic cancer 16–24 months before symptoms appear. This development could significantly increase survival chances in one of the most dangerous oncological diseases.
How Does Artificial Intelligence "See" the Invisible?
Unlike radiologists, who look for formed tumors, the artificial intelligence model developed by specialists at the Mayo Clinic and the University of Texas focuses on "precursors." During the training phase, the AI analyzed nearly a thousand tomograms to learn to distinguish the slightest anomalies in tissue density and architecture that do not yet appear as cancer but already indicate the onset of a pathological process.
What Are the Results of the Experiment?
During testing on a group of 63 patients, who were later confirmed to have cancer, the AI "sounded the alarm" in 46 cases long before the patients felt the first symptoms. In some cases, the system detected signs of the disease even two years before clinical confirmation. In comparison, on CT scans of healthy individuals, the AI almost did not produce false alarms.
Why Is the Disease Called the "Silent Killer"?
Pancreatic cancer does not give clear symptoms for a long time due to the location of the organ. The tumor is difficult to detect at an early stage, and signs appear only as the disease progresses.
Late symptoms include:
-
jaundice;
-
significant weight and appetite loss;
-
pain in the upper abdomen radiating to the back.
Why Is This Important?
To date, the survival rate for pancreatic cancer is one of the lowest among all cancers. The main reason is late-stage diagnosis when surgical intervention is no longer possible. If the model passes clinical trials on large groups of people, it could give doctors an additional 16–24 months of precious time to begin aggressive therapy or surgery.
AI is gradually becoming a tool for early diagnosis, capable of detecting disease where humans are currently powerless, writes bb.lv. If the technology proves effective in large-scale clinical trials, it could be a breakthrough in the fight against one of the most aggressive types of cancer.
