μCT-generated carpal cartilage surfaces: validation of a technique.

Computational models are increasingly being used for the analysis of kinematics and contact stresses in the wrist. To this point, however, the morphology of the carpal cartilage has been modeled simply, either with non-dimensional spring elements (in rigid body spring models) or via simple bone surface extrusions (e.g. for finite element models). In this work we describe an efficient method of generating high-resolution cartilage surfaces via micro-computed tomography (μCT) and registration to CT-generated bone surface models. The error associated with μCT imaging (at 10 μm) was 0.009 mm (95% confidence interval 0.007-0.012 mm ), or ~1.6% of the cartilage thickness. Registration error averaged 0.33±0.16 mm (97.5% confidence limit of ~0.55 mm in any one direction) and 2.42±1.56° (97.5% confidence limit of ~5.5° in any direction). The technique is immediately applicable to subject-specific models driven using kinematic data obtained through in vitro testing. However, the ultimate goal would be to generate a family of cartilage surfaces that could be scaled and/or morphed for application to models from live subjects and in vivo kinematic data.