Cookies on this website
We use cookies to ensure that we give you the best experience on our website. If you click 'Continue' we'll assume that you are happy to receive all cookies and you won't see this message again. Click 'Find out more' for information on how to change your cookie settings.

Research by Professor Jeremy Fairbank and colleagues, and published in the European Spine Journal, has found automation can improve how we diagnose back problems.

A novel feature of the model is to identify ‘Evidence Hotspots’ that are the array of elements that most contribute to the degradation scores. Jeremy Fairbanks

Back pain is the top cause of life-long disability worldwide, leading to enormous medical and social costs. Successful diagnosis of the problem at the root of the pain can lead to a better and more effective treatment, improving the lives of patients the world over.

The team looked at 12,018 intervertebral discs, from more than 2000 patients and concluded that the model developed to identify and label vertebrae and discs was as accurate as a radiologist at detecting disc degeneration. The model has the added advantages of being consistent (when compared to multiple practitioners), allowing for a faster analysis and therefore being cheaper.

This research has been recognised with the 2017 International Society for the Study of the Lumbar Spine Bioengineering Award.

"This new method opens the door to automated reading of Lumbar MRI scans. In a research setting this means large cohorts of scans can be read in a consistent fashion removing observer variation. In a clinical setting, where observer variation is also a significant problem, it provides a new method for quality control. We expect an early development will be to screen lumbar MRI scans for 'serious pathology', which is an important indication for the half a million lumbar MRI scans requested in NHS England each year. The impact of this will significantly reduce the cost of current scanning practice and to expand existing MRI capacity", said Professor Fairbank.

The project was based on a large cohort of subjects with back pain gathered and phenotyped as part of an EU funded study, Genodisc (The European Union Health Project on Genes and Disc Degeneration called 'Genodisc' - FP7 Health2007A Grant Agreement No. 201626).


Image: Examples of disc volumes (upper in each pair) and their corresponding evidence hotspots (lower in each pair). The leftmost and rightmost images are the second and eighth slice for each disc, out of the full volume of 9 slices. Note that these hotspots localise extremely well. These examples were randomly selected from different patients.