Asger Risom
A new AI model, Delphi-2M, can forecast health risks for more than 1,000 diseases, offering the potential to spot warning signs a decade before symptoms appear.
A new AI model, Delphi-2M, can forecast health risks for more than 1,000 diseases, offering the potential to spot warning signs a decade before symptoms appear.
A Breakthrough in Predictive Medicine
Researchers say Delphi-2M represents a major step toward personalized and preventative healthcare by modeling disease progression years in advance.
Built on Massive Datasets
The system was trained on data from 400,000 UK Biobank participants and tested on nearly 2 million patients from Denmark’s National Register.
Proof of Concept, Not Yet Clinical
While still experimental, scientists say the model already shows how AI can help identify health risks early and guide long-term care strategies.
Learning the “Grammar” of Health
Just as language models learn sentence structure, Delphi-2M analyzes health records as sequences of events—diagnoses, habits, and medical history.
Early Warnings for High-Risk Diseases
The AI performs especially well for conditions with clear biological patterns, such as certain cancers and cardiovascular disease.
Limits of Prediction
Accuracy is lower for illnesses shaped by unpredictable events, such as mental health disorders, highlighting the model’s current boundaries.
Risk Estimates Like Weather Forecasts
Delphi-2M doesn’t claim certainty but provides probability forecasts—much like predicting storms—helping doctors and patients prepare.
Towards Personalized Prevention
By identifying when risks emerge, the system could help shape early interventions, reducing the burden of late-stage disease treatment.
The Road Ahead
Researchers stress more validation is needed before clinical use, but the model opens a path to AI-powered disease prevention worldwide.
This article is made and published by Asger Risom, which may have used AI in the preparation
