🤖 An AI model capable of predicting over 1000 diseases and changes in health status decades into the future has become a breakthrough in medicine and personalized diagnostics. The development combined efforts from scientists at the European Molecular Biology Laboratory (EMBL), the German Cancer Center, and the University of Copenhagen. Algorithmic approaches similar to those underlying large language models—such as GPT or BERT—were applied, adapted for medical data.
The model was trained on huge datasets: anonymized data from 400,000 people in the UK Biobank study and 1.9 million patients from the Danish national registry. This allowed the AI to detect patterns in disease development that are not always obvious to humans.
According to Thomas Fitzgerald from the European Bioinformatics Institute, “Medical events often follow predictable patterns. Our model studies them and can predict future health outcomes.”
In other words, the AI can assess the probability of developing diseases such as cancer, diabetes, cardiovascular diseases, respiratory illnesses, and many other chronic and acute conditions.
Delphi-2M analyzes not only the patient’s medical history but also lifestyle factors: excess weight, habits, age, sex, which allows for highly personalized predictions. Risks are expressed as percentages over time, similar to a weather forecast: “for example, a 70% chance of developing the disease within the next five years.”
Acting EMBL Director Ewan Birney noted that in the coming years, patients will be able to gain practical benefits from such predictions: “You come to an appointment, and the doctor uses the tool and says, ‘Here are four main risks in your future, and here are two things you can do to change this.’” According to him, standard advice such as “lose weight” or “quit smoking” will remain relevant, but for some diseases, the model will provide more specific, scientifically grounded recommendations.
The key advantage of Delphi-2M is that it can simultaneously predict over 1000 diseases over a long-term horizon. “Prediction accuracy for each disease is comparable to existing specialized models while covering the full spectrum of potential risks,” the project team noted.
Professor Moritz Gerstung, head of AI at the German Cancer Center, emphasized that this is the beginning of a new way of understanding human health and disease progression. He noted that generative models in the future will be able to personalize medical care and anticipate healthcare system needs at the level of entire countries and regions.
It is worth noting that in September, researchers from Harvard Medical School presented an AI model capable of identifying precise gene and drug combinations to reverse pathological processes in human cells, further highlighting AI’s growing capabilities in medicine and pharmacology.
⚡ Thus, Delphi-2M and similar tools open a new era in medicine, where prediction, prevention, and personalized care become more accurate, scientifically grounded, and accessible to every patient. This could radically change approaches to treating chronic and rare diseases and help healthcare systems plan resources and preventive programs in advance.
All content provided on this website (https://wildinwest.com/) -including attachments, links, or referenced materials — is for informative and entertainment purposes only and should not be considered as financial advice. Third-party materials remain the property of their respective owners.