Novel imaging sequences and processing for detecting early musculoskeletal disease
Background
Osteoarthritis (OA) is a common condition, affecting between 16% and 18% of the population. Over a person’s lifetime, there is a 25% risk of developing late-stage OA. To avoid the pain and disability that comes with this, we need to identify and treat it earlier in order to prevent progression. Imaging is key to identifying OA, but current imaging methods are not sensitive enough to detect early changes. New Magnetic Resonance Imaging (MRI) techniques have strong potential, though further work is needed to identify OA accurately and to guide treatment.
What the research hopes to achieve
We want to detect the earliest changes in people’s hips and knees with the onset of OA. New MRI techniques, like T2 mapping (a technique to map different biological tissues on MRI), have the potential to do this, but how T2 relates to the tissue structure and what this means for OA is not yet fully understood.
To improve our understanding of these new imaging techniques, we need to be able to compare images from a large number of patients. Before we can compare patients, it is important to normalise images from different patients to a similar template (this is called image registration) so that they can be viewed and compared in the same ‘space’. Then we can generate T2 maps for different patients based on their MRI data, reduce these maps into a rectangular grid as shown above and look at the patterns revealed over a large number of patients. These patterns form a ‘fingerprint’ of disease which we can use to identify features of early OA and relate these to the best treatment option for the patient.
The outcome of this research will improve people’s understanding about early OA, and provide a way to detect early musculoskeletal disease before it is too late to be stopped.