How AI could personalise cancer care
Every cancer is different. Understanding those differences could help doctors diagnose cancer more accurately, choose the most effective treatment for each patient and reduce the need for some costly and time-consuming tests, making care more personalised and improving patients’ experience of treatment.
Barts Health is working with NHS, university and industry partners to explore how artificial intelligence (AI) and health data could make that possible.
The new DESTINY project will explore whether AI can create “digital twins” of human tissue, detailed virtual models that can help predict information about a patient’s cancer. Researchers hope this could reduce the need for some tests, while helping doctors understand how a cancer may behave and which treatments could work best.
Central to the project is the Barts Health Data Platform, which provides secure access to clinical data from one of the UK’s largest and most diverse patient populations. Linking this with tissue and molecular data held by Queen Mary University of London will allow researchers to test discoveries against real patient outcomes, helping to develop more accurate diagnoses and treatments that are more precisely targeted to individual patients.
The project will also explore how AI can learn securely from data held by different organisations without sensitive patient information being moved between them. This could make it possible to learn from larger and more diverse populations while protecting patient privacy.
Funded by Innovate UK, DESTINY brings together Barts Health, Guy’s and St Thomas’, King’s College London and Queen Mary University of London, alongside partners with expertise in AI and data science. It builds on the PharosAI collaboration and will initially test whether the approach can work at greater scale.
Fay Stevenson, Chief Information Officer at Barts Health, said: “This is about using the data we already hold securely to help improve cancer care. By connecting information about cancer, treatment and outcomes, we can help develop AI that could support more accurate diagnosis, more personalised treatment and potentially reduce the need for some tests.
“Our diverse patient population, clinical expertise and Data Platform put Barts Health in a strong position to turn advances in AI into real improvements in care and outcomes for patients.”
Read more