Estimation of 3D shape in the Patellofemoral Joint using Statistical Shape Models and 2D Data

Disorders of the patellofemoral joint (PFJ) including PF osteoarthritis and PF pain disorder have been estimated to represent 25% of all patients presenting for knee joint treatment. The diagnosis and treatment of these disorders is curtailed by lack of understanding of the mechanical operation of the joint. A crucial aspect to be considered in understanding contact patterns and kinematics of the PFJ is the alignment of the patella in the trochlear groove. Investigation of patellar alignment necessitates accurate knowledge of the 3D articular surfaces of both patella and femur, along with underlying bone geometry. These 3D data are generally obtained for bones and cartilage plates through reconstruction of images from computed tomography (CT) and/or magnetic resonance imaging (MRI) respectively. This research proposes to develop a parametric model to enable the accurate estimation of the patient-specific 3D shape of a patella and distal femur from several 2D images. There are several drivers for this work: the labour and cost (and radiation exposure in the case of CT) entailed in 3D scanning mean that it would be greatly advantageous to have the ability to characterise 3D joint geometry using 2D images obtained through the economical, timely and traditionally widely used x-ray / fluoroscopy method. Statistical shape modelling (SSM) and principal component analysis (PCA) has previously been applied to analysis of the distal femur with the aim of developing more sophisticated sizing and shaping rationales for replacement components. This level of shape analysis has not previously been applied to the patellofemoral joint