AI Model Trained to Predict Over 1,000 Diseases

AI tool forecasts over 1,000 diseases, predicting health changes a decade ahead.

Scientists have developed an AI tool capable of forecasting more than 1,000 diseases and predicting health changes up to a decade in advance.

Experts from the European Molecular Biology Laboratory (EMBL), the German Cancer Research Center, and the University of Copenhagen employed algorithmic principles similar to those used in large language models.

The AI was trained on data from two independent healthcare systems—anonymized information from 400,000 individuals in the UK Biobank study and 1.9 million patients from Denmark’s national registry.

“Medical events often follow predictable patterns. Our AI model learns these and can forecast future health outcomes,” stated Thomas Fitzgerald, a research fellow at the European Bioinformatics Institute.

The new tool assesses the likelihood of a person developing a particular disease and when it might occur. The neural network can predict cancer, diabetes, cardiovascular diseases, respiratory illnesses, and many other disorders.

The Delphi-2M model analyzes “medical events” in a patient’s history and lifestyle factors: excess weight, harmful habits, age, and gender.

Health risks are expressed as percentages over time, akin to a weather forecast: “70% chance of rain over the weekend.”

Acting Director of EMBL Ewan Birney mentioned that patients will begin to benefit from the tool in the coming years:

“You come to an appointment, and the doctor is already using such tools and says: ‘Here are the four main risks in your future, and here are two things you can do to change that.’”

He noted that standard advice like losing weight or quitting smoking will remain, but for some diseases, more specific recommendations will be available.

Birney highlighted that a key advantage of Delphi-2M over other solutions is its ability to predict all diseases simultaneously over a long period.

“Delphi-2M assesses the likelihood of more than 1,000 diseases considering individual medical history, and its accuracy is comparable to existing models for specific diseases,” the project team stated.

Head of AI at the German Cancer Research Center, Professor Moritz Gerstung, emphasized that Delphi-2M marks the beginning of a new way of understanding human health and disease progression. He stated that generative models could one day personalize care and anticipate healthcare needs across entire systems.

Earlier in September, researchers from Harvard Medical School introduced an AI model capable of identifying precise combinations of genes and drugs to reverse pathological states in human cells.