According to a study published in the journal Nature Medicine, An Artificial Intelligence (AI) tool can successfully identify people at risk for pancreatic cancer before three years from diagnosis. The researchers from the Harvard Medical school revealed that the diagnosis is solely based on the medical records of the patients. 

The researchers revealed that approximately 44% of people survive after the diagnosis if the cancer is detected in the early stages. However, only 12 % of the cases are diagnosed that early. They also noted that the rate of survival drops to 2 to 9% in patients whose tumors have grown beyond their site of origin. 

For the study, the researchers used two separate data sets totaling 9 million patient records from Denmark and the US. They asked the AI model to look for telltale signs based on the data contained in the records. Additionally, the researchers used multiple versions of AI model to identify individuals at increased risk of developing the cancer within different time frames, namely six months, a year, two years, and three years

The programme predicted which patients was likely to acquire pancreatic cancer in the future based on combinations of disease codes and the time of tumor growth. The researchers noted that some of the symptoms and disease codes were not directly linked to or caused by pancreatic dysfunction. 

Overall, each version of the AI algorithm was substantially more accurate at predicting the risk of pancreatic cancer in comparison to current population-wide estimates of disease incidence. 

(Source: https://indianexpress.com/article/technology/science/artificial-intelligence-pancreas-cancer-prediction-8600045/ )