New AI Model Forecasts Risk of Over 1,000 Diseases with High Accuracy
A new artificial intelligence tool, named Delphi-2M, is capable of predicting an individual’s risk of developing more than 1,000 diseases, including cancer, diabetes, heart disease, and respiratory illnesses, researchers announced today. Developed by scientists at the European Molecular Biology Laboratory’s European Bioinformatics Institute (EMBL-EBI), the model analyzes patient histories – encompassing diagnoses, lifestyle factors like obesity, smoking, and alcohol consumption, and also age and sex – too forecast potential health outcomes over the coming decade and beyond.
The breakthrough, detailed in a study published in the journal Nature, leverages data from two self-reliant healthcare systems: the UK Biobank study (400,000 individuals) and the Danish national patient registry (1.9 million patients). Delphi-2M expresses health risks as probabilities over time, similar to weather forecasting.
“Medical events frequently enough follow predictable patterns,” explained Tomas Fitzgerald, a staff scientist at EMBL-EBI. “Our AI model learns those patterns and can forecast future health outcomes.”
Unlike existing single-disease prediction models, Delphi-2M can assess risks across a broad spectrum of conditions simultaneously and over extended periods. “We can do all diseases at once and over a long time period. That is the thing that single disease models can’t do,” stated Ewan Birney, EMBL interim executive director.
Researchers anticipate the tool could be integrated into clinical practice within the next few years, enabling doctors to proactively identify and address patient risks. “You walk into the doctor’s surgery and the clinician is very used to using these tools, and they are able to say: ‘Here’s four major risks that are in your future and here’s two things you could do to really change that,'” Birney predicted.
The model’s “generative nature” also allows for the creation of synthetic future health trajectories, offering estimates of potential disease burden for up to 20 years. Prof Moritz Gerstung, head of the division of AI in oncology at the German Cancer research Center, hailed the progress as “the beginning of a new way to understand human health and disease progression,” suggesting it could ultimately led to personalized care and improved healthcare planning.
