Bringing academics and AI companies together for fracture diagnosis
The Digital Research Team is ideally placed at the interface between clinicians and commercial companies. One example of this is a project developing artificial intelligence (AI) technology to improve the diagnosis of limb fractures, bringing together usually inaccessible patient imaging data and AI start-ups.
Background
Fractures can be challenging to diagnose, particularly in emergency situations. In 2019 the Digital Research Team was involved in a project aiming to improve fracture diagnosis using artificial intelligence.
NHS Grampian had identified an issue with some patients needing to travel long distances from community hospitals or walk-in clinics to access accurate fracture diagnosis – often of the wrists and ankles – at city hospitals. Improving diagnosis closer to home could potentially save patients a journey, as well as reducing the workload of radiographers.
Complex cases weren't being diagnosed fast enough meaning that patients often had to be recalled and have more complex time-consuming procedures compared to if they had been identified earlier.
SHAIP
The Digital Research Team created SHAIP along with the University of Glasgow and Canon Medical Research Europe as a workstream of the Industrial Centre for Artificial Intelligence Research in Digital Diagnostics (iCAIRD).
SHAIP is the first ever data safe haven to hold and process image-based and free-text data, in this case in the form of upper limb (wrist or hand) and lower limb (ankle or foot) x-ray plates and associated clinicians’ notes.
The Project
The £240,000 project was a Small Business Research Initiative (SBRI) competition funded by Opportunity North East and NHS Scotland. It called for applicants with appropriate AI tools in the research and development phase to apply for funding to further develop and train the tools using data from the Safe Haven Artificial Intelligence Platform (SHAIP).
The project used innovative data technology to screen radiographs of wrist and ankle fractures to enable radiologists to focus on more complex cases.
“This project was one of the first to use AI on patient data in this way. It’s really exciting to see how this kind of tool of could improve patient care and free up clinicians’ time.”
“The Digital Research Team’s expertise was crucial to helping the companies navigate the regulatory pathways that ensure patient confidentiality, as well as give assurance to clinical partners.”
“The really innovative bit was the fact that ONE (private sector) and NHS co-funded this project bringing academics, clinicians, and industry together to collaborate on practical solutions.”
Five tech SMEs were chosen for the first phase, receiving £20,000 each to conduct a three-month feasibility study. Two organisations, Bering and See AI, then went on to a longer second phase where they each received £70,000 for prototype development and testing.
After training on both fractured and non-fractured x-ray plates, the tools were found to be effective at fracture diagnosis, achieving at least as good a diagnosis rate as clinicians. The companies are now in the process of pursuing regulatory approval for their tools to be put into clinical use.
Lorna Duguid, Life Sciences Director at Opportunity North East, says: “This project harnessed the expertise of clinicians and industry to produce better results for patients. Digital is transforming healthcare at an unprecedented pace and ONE’s co-funding for the project built on the region’s strengths in health data science, imaging, and analytics.
All the partners were encouraged by the results, and we look forward to seeing commercial solutions coming out of the collaborative innovation, demonstrating the region’s ability to drive digital health developments and bring forward solutions to healthcare challenges.”
“We want companies to know that we welcome them getting in touch with ideas about how SHAIP can be used, and the kinds of healthcare challenges that their tools could help with. We have the expertise to help companies that aren’t already in the healthcare sphere to transfer their knowledge – there could be some exciting collaborations in the future,” says Katie Wilde, the Digital Research Team’s Head of Digital Research.
For more information on working with SHAIP, please contact the Grampian Data Safe Haven at dash@abdn.ac.uk.
Built with Shorthand