The aim of this study was to discover if the ulnar styloid is sufficiently consistent in size, shape and position relative to other bony features of the ulna to be used as a reference in pre-operative planning of fixation of broken bones. The comparison of size and shape (together known as form) between bones has recently been facilitated thanks to the advance of technologies designed to allow the comparison of the form of structures using anatomical landmarks. This new class of methods is collectively known as geometric morphometrics. It eliminates the differences in location and rotation of landmark through registration that minimises the sum of squared deviations from each other after scaling. This is Procrustes registration. The residual size and shape information is amenable to statistical analysis. In the present application, the registered Procrustes landmarks are used to compute a mean (reference) shape. The individuals are then compared to this mean/ reference shape. Using principal components analysis (PCA) variations in shape are not only identi-fied, but also quantified. The identification of patterns of deviation from the mean shape is considerably enhanced through the use of 3-D visualistaions of the shape variations represented by the space of the PCA. These analyses indicate that the ulnar styloid is suf-ficiently consistent in location to other anatomical landmarks that it could be used as a radiographic marker in preoperative planning. More importantly, the analysis of this study indicates that the methods of geometric morphometrics are widely applicable to the analysis of 3-D variations in morphology facilitating the analysis and comparison of radiographs. A useful future application will be in the development of 3-D reference morphologies that will allow the surgeon to compare and contrast the morphology of a radiograph of a badly broken (comminuted) bone to a standard one. Eventually computer might assist the surgeon by geometrically and visually showing how and by how much the bone needs reduction. Similarly, applications to the the virtual comparison of diseased and healthy bones might allow quantative and visual comparisons that could aid diagnosis and planning.
