Aging adult skull vaults by applying the concept of fractal geometry to high-resolution computed tomography images

IntroductionAging human remains is a critical issue in anthropology and forensic medicine, and the search for accurate, new age-estimation methods is ongoing. In our study, we, therefore, explored a new approach to investigate a possible correlation between age-at-death (aad) and geometric irregularities in the bone structure of human skull caps. We applied the concept of fractal geometry and fractal dimension D analysis to describe heterogeneity within the bone structure.MethodsA high-resolution flat-panel computed tomography scanner (eXplore Locus Ultra) was used to obtain 229,500 images from 221 male and 120 female (total 341) European human skulls. Automated image analysis software was developed to evaluate the fractal dimension D, using the mass radius method. The frontal and the occipital portions of the skull caps of adult females and males were investigated separately. The age dependence of the fractal dimension D was studied by correlation analysis, and the prediction accuracy of age-at-death (aad) estimates for individual observations was calculated.ResultsD values for human skull caps scatter strongly as a function of age. We found sex-dependent correlation coefficients (CC) between D and age for adults (females CC = −0.67; males CC = −0.05). Prediction errors for aad estimates for individual observations were in the range of ±18 years at a 75% confidence interval.ConclusionsThe detailed quantitative description of age-dependent irregularities in the bone microarchitecture of skull vaults through fractal dimension analysis does not, as we had hoped, enable a new aging method. Severe scattering of the data leads to an estimation error that is too great for this method to be of practical relevance in aad estimates. Thus, we disclosed an interesting sex difference.