UK cancer centre launches AI-powered detection drive
Professor Dow-Mu Koh, professor of functional cancer imaging and a consultant radiologist at The Royal Marsden.
The Royal Marsden NHS Foundation Trust in the UK has joined forces with two private-sector companies on artificial intelligence models to detect cancer and guide treatment.
The internationally renowned cancer centre is working on the project – billed as the UK's first large-scale, AI-powered radiology platform for cancer research – with digital transformation specialist NTT Data Group and CARPL.ai, which focuses on applying AI to the interpretation of medical images.
The alliance revolves around an AI-powered radiology analysis service for the development and evaluation of AI in medical imaging, with the aim of improving detection rates, speeding up treatment, and ultimately improving patients' chances of survival.
It will focus on developing and evaluating AI algorithms to improve the accuracy of cancer evaluation, including sarcoma, lung, breast, brain and prostate cancers.
The Royal Marsden is home to the Radiology and Artificial Intelligence Research Hub, funded by The Royal Marsden Cancer Charity, which brings together clinicians and scientists for projects designed to benefit patients using radiology data. Ongoing projects involve the development of AIs to improve the diagnosis and treatment of sarcomas (RADSARC-R) and breast cancer (AI-BRACE).
AI's potential to enhance the speed and accuracy of identifying imaging biomarkers has long been recognised, but it is only in the last 10 to 15 years that advances in computing power and algorithm design have started to accelerate progress and there are still relatively few cases in which AI is being deployed routinely within the NHS.
According to the partners, the new service – which is being funded via a three-year grant from the UK's National Institute for Health and Care Research (NIHR) – will be used for research at The Royal Marsden Hospital and the Institute of Cancer Research (ICR). Once the service is live, research teams will be able to evaluate a range of AI models across several cancer types and explore how they can support clinical decision-making and shape future approaches to diagnosis and treatment.
The service runs on a machine learning operations (MLOps) clinical imaging platform, built and operated by NTT Data, which hosts a CARPL.ai platform that helps test and manage AI algorithms and already includes a large collection of radiology AIs.
"AI has immense potential to support clinicians in diagnosing and treating cancer earlier and more precisely," said Dow-Mu Koh, professor of functional cancer imaging and a consultant radiologist at The Royal Marsden.
"By working with NTT DATA and CARPL.ai, we've created a scalable research environment that allows us to explore the full potential of AI safely and in a way that could one day transform cancer diagnosis and treatment across the NHS," he added.
