Revolutionizing Medical Predictions: AI Forecasts Health Issues Like Weather

Tue Sep 30 2025 06:17:46 GMT+0300 (Eastern European Summer Time)

A new AI model, Delphi-2M, can predict health risks up to a decade in advance, mirroring weather forecasts in accuracy and function. Using data from over 400,000 patients, it aims to identify high-risk individuals to prevent diseases effectively.

Scientists have developed Delphi-2M, an AI model that uses patterns from medical records to forecast health risks over the next decade. By analyzing data from the UK Biobank and validating it against extensive Danish medical records, the model shows promising results in predicting conditions like heart disease and diabetes, aiming to facilitate early interventions.

Artificial intelligence can predict people's health problems over a decade into the future, say scientists.

The technology has learned to spot patterns in people's medical records to calculate their risk of more than 1,000 diseases.

The researchers envision using the AI model to identify high-risk patients, paving the way for preventive healthcare and insightful resource planning for hospitals years in advance.

Delphi-2M, the new AI model, utilizes similar technology to well-known AI chatbots. It analyzes anonymous medical data to forecast what health issues may arise in individuals.

This model does not predict an exact date for events such as heart attacks but evaluates the likelihood of developing certain diseases over time, akin to predicting a 70% chance of rain in a weather forecast.

Developed using extensive UK data regarding lifestyle and health records, Delphi-2M was subsequently validated with data from 1.9 million medical records in Denmark, showing good prediction accuracy.

Current applications may include spotting high-risk patients, providing tailored lifestyle recommendations, and informing healthcare resource allocation, aiming for improved patient care.

However, while the technology is promising, further refinement and testing are required before clinical implementation. The team aims to address biases in the existing model and expand its capabilities by incorporating additional medical data types.

Lead researcher Prof Ewan Birney expressed excitement about the model's potential, calling it the beginning of a new understanding of health and disease dynamics.

As research continues, it may transform how healthcare providers anticipate and respond to health needs on a large scale.