New AI Model Detects Pancreatic Cancer from Stool Samples with High Accuracy
Helsinki,Finland & Tehran,Iran – September 25,2024 – Researchers have developed an artificial intelligence model capable of accurately distinguishing between patients with pancreatic cancer and healthy individuals based solely on analysis of their gut bacteria. The groundbreaking study, conducted by teams from the Finnish University of Helsinki and the University of Shahid Bahtti Medical Sciences in Iran, was published in the journal Gut Pathogens and reported on by the British newspaper The Autonomous.
The research builds on the understanding that the human body hosts trillions of bacterial cells – exceeding the number of human cells by a ratio of approximately 40 to 30 trillion – and that these microbial communities play a critical role in health and disease.
Pancreatic cancer,frequently enough referred to as a “silent killer,” is notoriously difficult to detect early due to its often subtle symptoms. By the time symptoms manifest, the disease frequently has progressed to an advanced stage, limiting treatment options. The cancer often develops in the pancreatic duct, the tube connecting the pancreas to the small intestine. Tumors in this area can obstruct the flow of digestive enzymes, leading to metabolic issues and symptoms like chronic fatigue and weakness, which are easily overlooked or attributed to other causes.
Recognizing the potential of non-invasive diagnostics, the researchers focused on stool samples. Because pancreatic cancer typically originates in the intestinal portion of the pancreas, and given that most individuals have regular bowel movements, stool provides a readily accessible window into the body’s internal state.
The study involved collecting and analyzing stool samples to map the DNA of the bacteria present. Results revealed a significant difference in the gut bacterial composition of pancreatic cancer patients compared to healthy controls. Specifically, patients with pancreatic cancer exhibited reduced bacterial diversity, with alterations in the abundance of certain bacterial types.
Leveraging these findings, the research team created an AI model that can accurately differentiate between cancer patients and healthy individuals based on their gut bacterial profiles. The model’s performance suggests a potential for a novel, non-invasive screening tool for early cancer detection.
The study also highlighted the complex interplay between cancer and gut bacteria. Certain bacterial characteristics not only indicate the presence of the disease but are also altered *by* the disease itself, demonstrating a bidirectional relationship.