European scientists have developed a new artificial intelligence tool, Delphi-2M, that can predict personal risk of developing over 1,000 diseases and predict changes in health 10 years into the future.
European scientists have developed a new artificial intelligence tool, Delphi-2M, that can predict personal risk of developing over 1,000 diseases and predict changes in health 10 years into the future.
The Guardian reports that the generative AI tool was created by experts from the European Molecular Biology Laboratory (EMBL), the German Cancer Research Center and the University of Copenhagen using algorithmic concepts similar to those used in large language models. The AI was trained on data from two completely separate healthcare systems.
«Medical events often follow predictable patterns. Our AI model learns these patterns and can predict future health outcomes,» said Thomas Fitzgerald, a researcher at the European Bioinformatics Institute (EMBL).
The AI tool estimates the likelihood of someone developing diseases such as cancer, diabetes, heart disease, respiratory disease and many other disorders. Delphi-2M looks for «medical events» in a patient’s history, such as when the disease was diagnosed, as well as other important parameters: obesity, smoking or alcohol use, as well as age and gender.
The tool also analyzes anonymous data from patients' medical records to predict what might happen over the next decade and beyond.
The tool was trained and tested on anonymized patient data from 400,000 people participating in the UK Biobank study and 1.9 million patients from the Danish National Patient Registry. Health risks are expressed as time-series measures, similar to a rain forecast.
EMBL interim executive director Evan Birney predicts that patients will be able to use this tool within the next few years.
«You come to the doctor, and he, accustomed to using these tools, can say, ‘Here are four major risks that await you in the future, and here are two things you can do to change that,’» Birney describes working with Delphi-2M.
According to him, the advantage of the new artificial intelligence tool over existing methods of calculating health risks is that AI allows «to analyze all diseases simultaneously and over a long period of time.»
«Delphi-2M predicts the incidence of over 1,000 diseases, based on each individual’s disease history, with accuracy comparable to current models for individual diseases. The generative nature of Delphi-2M also allows for the sampling of synthetic future health trajectories, providing meaningful estimates of the potential disease burden for up to 20 years,» the developers explained.
Professor Moritz Gerstung, head of the Artificial Intelligence Department in Oncology at the German Cancer Research Center, believes that generative AI models such as Delphi-2M will be able to personalize medical care in the future and also help predict large-scale healthcare needs.
Last year, AI startup Calmara offered users a somewhat unusual AI-based service that involves taking photos of genitals. According to the developers, this should be done before sex to make sure that the partner does not have sexually transmitted diseases.
