Scientists have developed a new artificial intelligence tool that can predict the risk of developing over 1,000 diseases and forecast changes in health status up to a decade in advance.
- 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.
- Scientists used algorithmic concepts similar to those employed in large language models (LLM). The AI was trained on data from two healthcare systems. Specifically, the tool was trained on anonymized information from 400,000 patients from the UK Biobank study and data from 1.9 million people from the Danish national patient registry.
- The Delphi-2M system searches for “medical events” in medical history, considers lifestyle factors, as well as the patient’s age and gender. The tool also analyzes patient medical history data to predict what may happen over the next decade and beyond.
- Scientists believe that patients will be able to benefit from the new AI tool within the next few years. Based on working with such AI, a doctor could identify, for example, the four main risks for a patient and provide useful recommendations.
- According to specialists, the advantage of the new AI tool over existing ones (such as Qrisk for calculating heart attack risk over the next decade) is that it can predict all diseases simultaneously and for an extended period. Models focused on individual diseases cannot do this.
- The accuracy of Delphi-2M is comparable to that of AI created for predicting specific diseases.
