Detection of sexual dimorphism in the human neurocranium at local scale

In physical anthropology sexual dimorphism refers to the morphological differences observed in female and male individuals belonging to the same species. In the human cranium a number of anatomical traits are known to be sexual dimorphic. In this work, we present a geometric morphometric approach to automatically detect the most sexual dimorphic on skeletal collections. We applied the workflow on the human neurocranium and we defined, without an a priori definition of modules, which portions are most sexually dimorphic. We used a large sample of sex-known human 3D skulls to analyse the rate of sexual dimorphism found in the human neurocranium. We applied the Procrustes ANOVA on the best dimorphic patch found using the proposed workflow. We calculated the accuracy in discriminating sex in a sex-known sample by using our proposed model and the traditional approach