Delphi-2M Predicts Long-Term Individual Health Risks

A team of European biomedical researchers described Delphi-2M in a September 2025 Nature article, a generative AI that predicts long-term individual health trajectories using health records from over 400,000 UK Biobank participants. The model was validated on 1.9 million Danish hospital records spanning five decades and predicts risks for more than 1,000 diseases, potentially guiding targeted screening though it remains a research tool.
Key Points
- 1Trained Delphi-2M on UK Biobank data from over 400,000 participants to predict health trajectories
- 2Validated using 1.9 million Danish hospital records across five decades, predicting over 1,000 diseases
- 3Offers clinicians simultaneous multi-disease risk forecasts, enabling targeted screening and preventive interventions
Scoring Rationale
High novelty and peer-reviewed validation, but still research-stage with limited immediate clinical deployment and generalizability concerns.
Sources
Public references used for this report.
Practice with real Banking data
90 SQL & Python problems · 15 industry datasets
250 free problems · No credit card
See all Banking problems

