Skip to main content

This site is best viewed with a modern browser. You appear to be using an old version of Internet Explorer.

Using Artificial Intelligence to lead the way in fracture prevention

24/06/2019
This article is more than four years old.

The Fracture Prevention Service at Oxford University Hospitals (OUH) has become the first unit in the United Kingdom to identify a bone weakening condition by using AI technology.

By using a software programme produced by company Zebra Medical, our team is able to detect osteoporotic vertebral fractures from all CT scanners in our hospitals.

Osteoporotic vertebral fractures occur when the struts that make up the mesh-like structure within bones become fragile and break easily following a minor injury or action such as heavy lifting.

Pioneered by a nurse-led quality improvement project, this technology has resulted in significant advances in care for patients with osteoporosis, a condition which can often lead to a reduced quality of life. Benefits include earlier detection of these fractures, which are often undiagnosed as patients aren't even aware they have them.

Sarah Connacher, a Specialist Nurse Practitioner in Fracture Prevention and Osteoporosis at Oxford University Hospitals NHS Foundation Trust, said: "Using AI technology to detect vertebral fractures has been an exciting development to the already excellent Fracture Prevention Service in Oxfordshire.

"This is something to be celebrated, not only for it being a nurse-led service but also for the significant advancement in care for vulnerable patients."

While the Oxford Fracture Prevention Service ensures all patients presenting with a fragility fracture (a fracture sustained following a fall from standing height) to our trauma services or Emergency Departments are seen and assessed for their osteoporosis risk, patients with an osteoporotic vertebral fracture are often missed as they rarely present to the hospital and remain undiagnosed.

Using AI has helped the Oxford Fracture Liaison Service (FLS), based at the Nuffield Orthopaedic Centre in Oxford, with its aim of identifying all patients with an osteoporotic vertebral fracture.

In addition to improving patient care, early identification and treatment of these fractures will result in savings for the NHS as a vertebral fracture can predict future fragility fractures, such as fractured hips and wrists.

The AI algorithm re-analyses the CT images of any OUH patient over the age of 50. The CT scans will have been carried out for reasons other than a vertebral fracture, meaning the software can identify vertebral fractures incidentally.

These scans are screened by the AI software and those positive for fracture are identified and analysed further by experienced nurses.

So far, more than 100 patients with a previously unknown osteoporotic vertebral fracture have been successfully identified, assessed, and put onto osteoporosis treatment by the FLS team - thanks to the help of AI.

The FLS team presented their work at the World Congress in Osteoporosis in Paris in April 2019. Zebra Medical, which sponsored the nurses to attend the conference, presented the nurses and Dr Kassim Javaid, an OUH Rheumatologist, with an award as a thank you and to highlight the excellent work they have achieved.