Predict Normal Shape of Femur for Cam-type Deformity Using Statistical Shape Model

Cam-effect impingement, which is related to femoral headneck deformity, is a common cause of hip impingement. Statistical shape modelling technique which aims to calculate a mean shape and to extract mathematical descriptions of morphological variations from a set of complex shapes was used to predict a normal shape for a cam-type femur based on landmarks taken from normal area of this cam-type femur. A statistical shape model (SSM) was constructed from a set of normal femora cut below the lesser trochanter. The cam-type femur was also cut below the lesser trochanter and registered to the common coordinate system of the SSM. Landmarks were taken from those normal areas of the cam-type femur. A new shape fitting the landmark cloud was then constructed by changing the morphological variations from the mean model iteratively until the root mean square (RMS) distance between the resulting shape and the landmark cloud was smaller than 0.05mm. The comparison between the prediction and the original model shows significant differences at the deformity area, with a maximum distance of 5.5mm. This indicates that this type of modelling can be used to provide a normal prediction for bones with deformity.