Overview | Publications | Research | Projects | Software | Miscellany

Projects on Musculoskeletal Image Interpretation

We have over 20 years experience in medical image analysis, with a focus on using statistical models of shape and appearance to extract information about bones.
In particular we have modelled their shape using Point Distribution Models and appearance with Combined Appearance Models.
We have located features with Active Shape Models, Active Appearance Models and Constrained Local Models.
We have developed Groupwise Image Registration to find accurate correspondences across large sets of images.
Here are some of our recent projects in the area.
Automated Analysis of Structures in 2D Medical Images
With Claudia Lindner
See also the www.bone-finder.com website.
Fully automatic cephalometric analysis
With Claudia Lindner
A tool to identify vertebral fractures
Studying Osteoarthritis in the knee.
ROAM Project
With Andrew Gait
(Funded by Arthritis Research UK)
Tools to analyse radiographs of the knee
(with Jessie Thompson)
A Statistical Shape Analysis of the Proximal Femur
with Claudia Lindner, (co-supervised by Gillian Wallis)
Wrist shape analysis
(with Raja Ebsim)
Analysing Lateral Knee Radiographs
(with Luca Minciullo)
The analysis of vertebral structure
(Funded by Arthritis Research UK)
Methods for groupwise registration of complex structures
With Pei Zhang
STOpFrac: Software for Opportunistic diagnosis of vertebral Fractures
Funded by NIHR
With Paul Bromiley
Estimating Skeletal Maturity from hand radiographs
With Steve Adeshina