Researchers pilot AI model to predict future health outcomes
De-identified NHS data from 57 million people in England is being used to train a generative AI model capable of predicting future health outcomes.
The model, called Foresight, is being tested by researchers at University College London (UCL) and King’s College London (KCL) as part of a national pilot exploring how AI could support earlier intervention and more personalised care.
Researchers hope the model will be able to make accurate predictions across demographic groups, including for people with rare conditions, help surface patterns that point to health inequalities and support service planning through population-level risk analysis.
In a press release, lead researcher Chris Tomlinson, from UCL Institute of Health Informatics, said: “AI models are only as good as the data on which they’re trained.
“So if we want a model that can benefit all patients, with all conditions, then the AI needs to have seen that during training.
“Using national-scale data allows us to represent the kaleidoscopic diversity of England’s population, particularly for minority groups and rare diseases, which are often excluded from research.”
Foresight functions similarly to large language models like ChatGPT, which are trained to recognise patterns in large datasets and predict what comes next in a sequence of data.
However, instead of predicting text, Foresight draws on routinely collected information such as hospital admissions, Covid-19 vaccination records and A&E attendances to forecast health events such as hospitalisations or new diagnoses.
A KCL-led study in March 2024 found that Foresight could accurately forecast a range of future medical events using NHS electronic health records, including new diagnoses, symptoms and medications.
The new pilot is currently restricted to Covid-19 research, using a limited number of datasets from between November 2018 and the end of 2023.
Richard Dobson, a lead researcher on the project and deputy director of the NIHR Maudsley Biomedical Research Centre, said: “Currently the data in this pilot is broad but shallow, and ultimately we’d like to harness the expertise and AI platforms behind Foresight by including richer sources of information like clinicians’ notes, or results of investigations such as blood tests and scans if they become available.”
Foresight operates entirely within the NHS England Secure Data Environment (SDE), allowing researchers to access de-identified health data while ensuring that both the data and AI model remain under NHS control.
Vin Diwakar, national director of transformation at NHS England, said: “AI has the potential to transform the way we prevent and treat disease, if trained on large datasets and safely tested.
“The NHS SDE has been fundamental to this pioneering research, shaping a future where earlier treatments and interventions are targeted to those who will benefit, preventing future ill health.”
Wes Streeting, health secretary, said: “I’m determined that we use this kind of groundbreaking technology to cut down on unnecessary hospital trips, speed up diagnosis times, and free up staff time.”
The pilot is supported by partners including Health Data Research UK, the National Institute for Health and Care Research and the British Heart Foundation Data Science Centre.
