For identifying shape changes in the human mandible associated with age, two factors were considered: shape changes associated with tooth-loss and general aging. Using CT-scans, automated landmark and surface mesh extraction of the mandible was performed on 115 individuals from Germany and China. As the Chinese individuals were only little affected by tooth-loss, even at very old age, they were used as control set for age-related changes not attributed to tooth-loss. The German sample, however, showed severe tooth-loss for older individuals. The landmarks were used to establish an initial spatial orientation within the sample. Based on this information surface registration was performed by applying an elastic ICP-Algorithm consisting of two parts: An iterative Gaussian smoothed deformation is performed with the resulting deformation being additionally regularized by minimizing deformation energy. Due to ambivalent quality of the CT-data in the condylar region and the coronoid processes, as well as the inconsistent presence/absence of teeth, these areas were excluded from further statistical analyses. The resulting set of registered coordinates was statistically evaluated for effects of aging. The predictor age provided a high explanatory value within the German sub-sample and a low one in the Chinese data. Visualizations show a similar pattern in both populations in the gonial region. But, while these are the only shape differences among Chinese (not affected by tooth-loss), Germans exhibit significant “shrinkage” of the corpus mandibulae, when compared to the rami, owed to atrophication caused by tooth-loss.
#changesinmandible #mandible
