US EditionNew AI model could predict risk of disease a decade before symptoms
- Researchers created an AI system called Delphi-2M that analyzes extensive medical records to estimate an individual’s risk and likely onset period for more than a thousand different health conditions, with predictions extending up to two decades into the future.
- This model was developed through a joint effort involving leading research institutions in Europe, including centers specializing in molecular biology, cancer research, and academia, to tackle the challenges posed by increasing chronic diseases and aging demographics.
- Delphi-2M was trained on anonymized medical data from 400,000 UK Biobank participants and tested on 1.9 million Danish patients, showing well-calibrated risk estimates but still needing further testing before clinical use.
- Prof Ewan Birney described their AI model as a demonstration that AI can learn long-term health patterns to make meaningful predictions, while Prof Moritz Gerstung highlighted that this approach represents an important advancement in exploring how diseases develop and progress, potentially enabling more personalized and preventative healthcare strategies.
- The model's forecasts could enable earlier personalized interventions and healthcare planning, though psychological effects on patients and ethical considerations require careful assessment before adoption.
22 Articles
22 Articles
Scientists Develop AI Tool That Can Predict Diseases Years In Advance
Scientists in Europe have unveiled an artificial intelligence tool capable of forecasting health risks years in advance. The model, called Delphi-2M, analyzes patterns in medical records to estimate the likelihood of more than 1,200 diseases, much like a weather forecast predicting a chance of rain. Developed with data from the UK Biobank and tested on nearly two million patient records in Denmark, the tool is especially accurate for conditions …
Estimating how human health changes over time, predicting the risk of developing over 1,000 different diseases, answering the question of which diseases will be met in the next 20 years. It is the perspective offered by the new model of Artificial Intelligence called Delphi-2M, obtained with the research conducted by the European Laboratory of Molecular Biology (Embl) and the German Center for Cancer Research (Dkfz), which trained the model with…
Scientists train AI model to predict future illnesses
PARIS, France — Scientists said Wednesday that they had created an AI model able to predict medical diagnoses years in advance, building on the same technology behind consumer chatbots like ChatGPT. Based on a patient’s case history, the Delphi-2M AI “predicts the rates of more than 1,000 diseases” years into the future, the team from British, Danish, German and Swiss institutions wrote in a paper published in the journal Nature. Researchers tra…
Trained with data from 400,000 Britons and validated in almost 2 million Danes, the tool, presented in a ‘Nature’ study, can anticipate risks of more than 1,000 pathologies and simulate health evolution over two decades Read
Coverage Details
Bias Distribution
- 40% of the sources lean Left
Factuality
To view factuality data please Upgrade to Premium
