The use of artificial intelligence in medicine is rapidly expanding. A study published September 17 in Nature details the creation of a generative AI by European researchers that could revolutionize the medical field. This AI, named Delphi-2M, can detect the risk of developing certain diseases, predicting their onset more than a decade before diagnosis.

Researchers from the European molecular biology laboratory, the German Center for Research on Cancer, and the University of Copenhagen developed Delphi-2M using data from two distinct healthcare systems. It’s creators believe it is a compelling demonstration of AI’s potential in medicine, particularly in modeling the advancement of human diseases.

Large-Scale Predictive Capabilities

Delphi-2M has the potential to detect the risk of over 1,000 diseases, including cancer and diabetes, once it becomes widely available to medical professionals and patients. However,its reliability decreases for diseases with more random development patterns,such as infections or mental health disorders. The tool primarily assesses the likelihood of a person developing heart or respiratory diseases, cancer, or diabetes.

According to The Guardian, a “prediction” from Delphi-2M indicates a risk that patients can proactively address, not a guaranteed outcome.Delphi-2M was trained and tested using data from over 400,000 anonymous patients in the UK BioBank study and 1.9 million patients in the danish national patient register. AFP, as cited by TF1, reports that the AI will undergo further testing before broader use. according to Ewan Birney, acting executive director of the European molecular biology laboratory, patients could benefit from this tool within the next few years.

“A prediction will therefore not mean a certain development of any pathology, only a risk that the patient can thus prevent.”