Cambridge researchers win government funding for their artificial intelligence (AI) technologies

Technologies developed by Cambridge researchers that use artificial intelligence to speed up diagnosis and improve patient care have been successful in the latest round of the £140million Artificial Intelligence in Health and Care Award.

In total four AI projects, which Cambridge researchers either led or collaborated in, received funding, which was announced by Secretary of State for Health, Matt Hancock, on 16 June.

It means that the researchers will be able to take the technology one step closer to being used in the NHS to benefit patients.

AI-systems for improving blood transfusion outcomes

Professor Emanuele Di Angelantonio from the NIHR Blood and Transplant Research Unit in Donor Health and Genomics, Dr William Astle NHSBT Senior Lecturer in Statistical Science in the MRC Biostatistics Unit and their collaborators at UCLH, Oxford University and international blood services, have been awarded more than £1million to improve, develop and implement AI-systems for genetic blood group typing, the automated stocking of blood according to type and the precision matching of patients to blood units.

Professor di Angelantonio said: “There are over thirty known blood group systems and each system can create different blood types.
“For patients receiving regular blood transfusions, it’s vital that the blood they receive is compatible with their own. If the blood is not matched, it can cause complications which can get worse over time – but at the moment it costs the NHS a lot of time and money to run the tests that are needed to measure a complete set of blood groups.

“It is now very cheap to measure blood groups genetically – we hope to blood type 100,000 blood donors and 500 sickle cell disease patients genetically in 2021. This will allow us to use AI to match sickle cell disease patients – who stand to benefit hugely from more precise blood matching – to donors.”

Dr Astle added: “This NIHR AI Award funding will help us to extend and develop the AI methods required for genetic blood matching, which should reduce the avoidable harm caused by transfusion reactions in the NHS.
“This has the potential to transform the quality of clinical care for patients dependent on transfusion, who can become untransfusable if they regularly receive mismatched blood.”

AI-enabled spine fracture pathway

In this multi-centre study, five hospitals including Cambridge University Hospitals are using an AI solution to detect osteoporotic fractures and identify new patients for treatment.

Dr Ken Poole, who is leading the study in Cambridge, said: “Osteoporosis is a common bone disease. Breaking a bone in the spine is a clear sign of it, but many patients don’t even realise they have a spine fracture at the time. They mistake the symptoms for ordinary back pain, and ignoring that cue can then lead to a cascade of more spine fractures or even a hip fracture, with devastating consequences.

“Each year over two million people undergo scans that include the spine, for various reasons such as lung or bowel problems. Remarkably, up to one in 20 of people having these scans could have spine fracture, although very few fractures are recognised or acted upon.

“This project will use an innovative AI software that automatically looks at existing CT scans to find these fractures and brings them directly to the specialist team’s attention, to see if the patient needs bone-strengthening lifestyle advice and medicines.

“We believe that this ‘AI-enabled spine fracture pathway’ will improve patient health and reduce costs to hospitals – and ensure tens of thousands of adults with undetected spine fractures are identified and protected against having future potentially life-changing fractures.”

This project is being run with Zebra-Med, an AI and machine learning company headquartered in Israel.

AI to differentiate tumour and healthy tissue on cancer scans

Consultant oncologist Dr Raj Jena has received an AI Award for 12 months to accelerate the process of registering AI technology that can spot the difference between tumours and healthy tissue on cancer scans as a medical device.

Once registered, it can then begin to be rolled out across key NIHR BRCs.

The NIHR Cambridge and Birmingham BRCs are using open-source AI tools from Microsoft Project InnerEye to differentiate tumour from healthy tissue on cancer scans. This is called ‘segmenting’ and takes place prior to radiotherapy treatment.

Dr Jena said: “This saves clinicians’ time, and reduces the time between the scan and commencing radiotherapy treatment.”

Monitoring slow-growing brain tumours

Certain types of brain tumour are deemed low-risk, as they grow so slowly. This project, led by Hon. Consultant Neurosurgeon Stephen Price, aims to develop AI to measure the volume of tumours from scans, and learn which are at risk of growth, to ensure those patients are monitored more frequently, and others can be reassured that their tumour is lower risk.

Professor Miles Parkes, Director of the NIHR Cambridge BRC, said: “This is fantastic news and congratulations to everyone involved.

“The Award will help increase the impact of the AI-driven technologies that Ken, Emanuele, Will, Raj and Stephen are developing.

“Their innovations will not only offer faster and more personalised diagnosis, they will also provide the evidence we need to demonstrate the effectiveness and safety of AI-driven technologies in health and social care.

“The technologies our researchers are developing today have the potential to transform clinical care for thousands of patients across the country.”

Patient and Public Involvement and Engagement (PPI/E) Strategy Lead Dr Amanda Stranks said: “For these technologies to be successful in the clinic, it is essential that patients are at the heart of their design and usage.

“It’s also great to see the NIHR asking for clear evidence of strong patient and public involvement strategies in the projects they are funding.

“For example, the AI osteoporosis study under Ken Poole in Cambridge will have patients represented on the study steering committee, project management team and locally, to ensure the project remains patient-centred throughout its life cycle.

“The AI system for improving blood transfusion outcomes also has actively involved patients, including setting up a patient panel to advise on the project.”

• All the AI projects that receive awards will be independently evaluated for their effectiveness, safety and value in the NHS settings in which they are deployed. This will add to the evidence base and inform the onward adoption and scaling of the technology.