Linking imaging and clinical data to improve orthopaedic surgery

Orthopaedic surgery often changes how a joint moves, sometimes in ways that are intended and sometimes in ways that are not. Even small differences in a person’s anatomy can affect how well a joint functions after surgery.

Because current tools for predicting these biomechanical changes are limited, some patients may not get the best possible outcome. Examples include knee and hip replacements, ligament repairs, osteotomies (bone realignment) and procedures to stabilise joints. 

New technology, especially machine learning and computer-vision methods, can help by identifying key features on scans and building detailed models of a person’s bones and joints. These models can move beyond simple 2D images to create dynamic 3D and even “4D” models showing joint motion.  

Early pilot work has shown this is possible, but current methods are slow and require a lot of manual work. To make these tools usable at scale in the NHS, they need to be trained on large, high-quality datasets that include both clinical images and patient information. 

Project aims

This project paves the way for secure access of orthopaedic imaging (x-rays, MRI, CT) and linked electronic patient record data, within the South West Secure Data Environment (SDE).  

We want to create reliable processes for data access, linkage and analysis. This will support the development, training and testing of advanced tools for recognising anatomical features, segmenting images and modelling joint shape and movement.

We will also incorporate methods to extract useful information from clinical letters using natural language processing. 

What we hope to achieve

By bringing these datasets together and improving the technical pipelines needed to analyse them, we aim to accelerate the development of semi-automated and automated tools for planning orthopaedic surgery.  

This will be a key step toward using personalised, computer-assisted modelling in routine NHS care. The work will also form the foundation for future large national research grant applications, helping move the technology closer to everyday clinical use.