AI can provide for your future health – just like the weather

Artificial intelligence can predict people’s health problems for a decade in the future, scientists say.

Technology has learned to identify models in people’s medical records to calculate their risk of more than 1,000 diseases.

Researchers say that it is like a weather forecast that anticipates 70% of rain – but for human health.

Their vision is to use the AI ​​model to identify patients at high risk to prevent diseases and help hospitals understand demand in their region, years in advance.

The model – called Delphi -2M – uses technology similar to well -known IA chatbots as chatgpt.

IA chatbots are formed to understand language models so that they can predict the sequence of words in a sentence.

Delphi-2M was trained to find models in anonymous medical records so that he can predict what comes and when.

He does not predict the exact dates, as a heart attack on October 1, but rather estimates the probability of 1,231 diseases.

“So, just like the time, where we could have 70% of the rain, we can do it for health care,” said Professor Ewan Birney, acting director of the European Molecular Biology Laboratory.

“And we can do it not only for a disease, but on all diseases at the same time-we have never been able to do it before. I am excited,” he said.

The AI ​​model was initially developed using data in the United Kingdom Anonymous – including admissions to the hospital, GP files and lifestyle habits such as smoking – collected from more than 400,000 people as part of the British biobank research project.

The model was then tested to see if its predictions stacked using data from other participants in the Biobank, then with 1.9 million people’s medical records in Denmark.

“It’s good, it’s really good in Denmark,” said Professor Birney.

“If our model says that it is a risk of 10 for next year, it really seems that it turns out to be one in 10.”

The model is the best to predict diseases such as type 2 diabetes, heart attacks and sepsis that have a clear progression of the disease, rather than more random events such as infections.

People are already offered a hypocholesterolemic statine according to a calculation of their risk of heart attack or stroke.

The AI ​​tool is not ready for clinical use, but the plan is to use it in a similar way, to identify patients at high risk when it is possible to intervene early and to prevent the disease.

This could include specific drugs or lifestyle advice – such as people likely to develop certain liver disorders benefiting from the reduction of their alcohol consumption more than the general population.

Artificial intelligence could also help to shed light on disease screening programs and analyze all health care files in a field to anticipate demand – such as the number of heart attacks per year in Norwich in 2030, to help plan resources.

“This is the beginning of a new way of understanding the progression of human health and disease,” said Professor Moritz Gerstung, head of the AI ​​division in oncology in DKFZ, the German cancer research center.

He added: “Generative models such as ours could one day help personalize care and anticipate large -scale health care needs.”

The AI ​​model, described in the life scientific journal, needs refining and test before its clinically use.

There are also potential biases because they have been built from British Biobanc data which is mainly drawn from people aged 40 to 70, rather than the whole population.

The model is now upgraded to take into account more medical data such as imaging, genetics and blood analysis.

But Professor Birney says: “Just to emphasize, it is research – everything must be tested and well regulated and thoughtful before its use, but technology is there to do this kind of prediction.”

It provides that this will follow a path similar to the use of genomics in health care where it took a decade to pass confident scientists in health care technology able to use it regularly.

The study was a collaboration between the European Molecular Biology Laboratory, the German Cancer Research Center (DKFZ) and the University of Copenhagen.

Professor Gustavo Sude, researcher in neuroimagery and AI at King’s College London, said: “This research seems to be an important step towards evolution, interpretable and – especially – ethically responsible for predictive modeling in medicine.”