Researchers say a new AI model can estimate a person’s risk of more than 1,000 diseases over time. It forecasts when illnesses like cancer or heart attacks may occur, and how risks change across years.
According to the Mirror, the study appears in Nature and reports results from large UK and Danish datasets.
How the Delphi-2M model learns risk
The team trained the system on anonymised records from 400,000 people in the UK Biobank. They then tested it with data from 1.9 million patients in the Danish National Patient Registry.
The model studies medical events in each person’s history. It tracks diagnoses and the time between them, and also uses age, sex, obesity, smoking, and alcohol use. It then predicts if and when future conditions may develop.
Health risks are expressed as rates over time, like a weather forecast with a chance of rain. The tool performed better for diseases with clear progression patterns, including some cancers and heart attacks. It was less reliable for variable conditions such as mental health problems or pregnancy complications.
What researchers and clinicians say
Ewan Birney of the European Molecular Biology Laboratory said clinicians could use such tools in five to 10 years. He described a visit where a doctor lists major future risks and actions to take.
Moritz Gerstung, who leads the division of AI in oncology at the German Cancer Research Centre, called it a new way to read disease progression. He said generative models could help personalise care and scale planning.
Study claims and next steps
The paper states: “Delphi-2M predicts the rates of more than 1,000 diseases, conditional on each individual’s past disease history, with accuracy comparable to that of existing single-disease models.” It adds the model can sample synthetic future health paths for up to 20 years.
The study reports that predictions were most useful when disease courses are well defined. The team says the approach can estimate potential disease burden over long periods.
Birney suggested common advice such as weight loss and smoking cessation will remain. He said some diseases may also see very specific guidance in future. The researchers present the tool as support for earlier and more tailored care.
The Mirror reports that scientists see the AI as an aid for doctors, not a replacement. They expect progress as more data, validation, and clinical use develop over the next decade.
