Development of morphological filtering method for specification and characterisation of hip replacement taper junctions

The introduction of modular large head metal-on-metal (LHMoM) hip replacements promised lower wear rates and reduced chance of dislocation coupled with an increased range of motion compared to conventional metal-on-metal hips. However, the 2010 National Joint Registry reports a 5 year failure rate of LHMoM hips of 7.8% compared to 6.3% for hip resurfacings and comparisons of well functioning hips have shown the LHMoM have significantly increased metal ion levels compared to resurfacings. This has led to a widely held hypothesis that the modular head/neck taper junction could be a site of secondary wear debris and indeed wear has been indentified in this area on a number of retrieved bearing tapers.
It is clear that the correct specification of this junction is vital to the optimum in-vivo performance of these bearings and that this could reduce the number of observed failures. This study will investigate the example of a number of newly manufactured total hip replacement femoral stems and will be the first to develop a method for specifying and measuring the underlying form of the textured surface of the critical neck taper which is characterised by a micro-thread. This was achieved through developing a filtering routine that allows for the extraction and isolation of the profile points that would contact the femoral head female taper counterface when in operation.
In this study a series of measurements were taken on each component, axially along the neck taper. The measurements were performed using a Talyrond 365 roundness machine with a 5m diamond stylus. The subsequently extracted profiles were firstly levelled using a least squares method in a post-process routine and were then processed using a newly developed morphological filtering routine to extract the envelope profile of contact points from the primary texture inherent in each profile.
The proposed morphological filtering method differs from the traditional implementation in that instead of calculating the extreme value in height for each sampling position the algorithm sets out to compute the contact points as the ball is rolling over the profile. The proposed method is derived from the Graham scan algorithm original developed for computing the convex hull of the planar point set. The algorithm handles the points in an incremental manner and a stack structure is maintained to hold the contact points by evaluating the coming point with those in the stack. To correct the end effect of filtration on the open profile, two ends of the profile are padded by reflection before the filtering process. At the final of computation, the stack contains all the contact points and the envelope ordinates are archived by interpolating points on the arcs determined by the adjacent contact point pairs.
The application of the unique morphological filtering method allows for the relationship between stylus ball size and the consequent measurement result to be investigated through application of different morphological filter sizes. The use of the morphological filter in this way allows for the production of a statement of the optimised measurement parameters for this critical and yet previously overlooked feature.