Researchers have developed AI that can predict future diseases

Date:

Share this article:

Del denne artikel:

A new AI system may soon predict a person’s future health decades in advance. It's raising hopes for prevention, but also questions about trust and data.

Artificial intelligence has rapidly moved from the realm of science fiction into hospitals and clinics.

Since its public breakthrough in 2022, the technology has reshaped industries from finance to education, and now, it’s transforming healthcare.

According to a new study published in Nature and based on research from the UK Biobank, scientists have developed an AI system called Delphi-2M that may forecast an individual’s risk of developing certain diseases long before symptoms appear.

How Delphi-2M works

Unlike traditional diagnostic tools, Delphi-2M doesn’t focus on a single illness.

Also read: Experts explain why Botox seems less effective when you use Wegovy

Instead, it analyzes massive health datasets, including medical histories, age, lifestyle habits, and clinical records, to identify patterns linked to more than a thousand conditions.

Trained on information from over 400,000 UK participants and tested on nearly two million people in Denmark, the model reportedly predicted health outcomes with remarkable precision.

Researchers found its mortality forecasts reached a score of 0.97, where 1 would indicate perfect accuracy.

Still, experts caution that such results depend heavily on the data’s quality.

Also read: Study shows nearly 60 % of meditators experience unwanted side effects

As scientists often warn about the results. If the datasets contain biases, for example, underrepresenting older or less healthy populations, the predictions may also reflect those gaps.

Promise and caution ahead

Doctors and data scientists see huge potential for preventive medicine.

Tools like Delphi-2M could help anticipate age-related diseases and guide early interventions long before illness develops.

Yet, the technology also raises ethical questions about privacy, consent, and fairness.

Also read: The most beautiful boys’ names in the world, according to science

Sources: El Confidencial, and Nature.

Also read: How to avoid food poisoning from leftover rice, potatoes and pasta

This article is made and published by Ida-Marie Palm Varbæk, who may have used AI in the preparation.

Other articles

An AI learned to see a migraine coming the next day, from a signal clinicians often overlook

Built on nearly 800,000 diary days from people living with migraine, the forecast was far more accurate than many had expected.

The same workout may affect men’s and women’s livers differently

Researchers watched the same workout play out inside the body in two very different ways, depending on sex.

Why fruit and vegetables are placed at the entrance of supermarkets

Product placement in supermarkets is no coincidence. Stores use various methods to influence what customers buy and how much money they spend.

Want to live longer? These foods may be worth eating more of

What do people who live to 100 or older eat? Researchers have studied dietary habits in regions where...

An AI learned to see a migraine coming the next day, from a signal clinicians often overlook

Built on nearly 800,000 diary days from people living with migraine, the forecast was far more accurate than many had expected.

The same workout may affect men’s and women’s livers differently

Researchers watched the same workout play out inside the body in two very different ways, depending on sex.

Why fruit and vegetables are placed at the entrance of supermarkets

Product placement in supermarkets is no coincidence. Stores use various methods to influence what customers buy and how much money they spend.