Geometric morphometric and traditional methods for sex assessment using the posterior ilium

The human hip bone is generally accepted as the most reliable bone for sex estimation in forensic and bioarchaeological disciplines. However, it is seldom completely preserved. The best preserved region is typically around the sacroiliac joint and its auricular surface; it is therefore surprising that this surface has not been involved in standard sexing methods. The aim of this study was to explore the shape and size sexual dimorphism of the auricular surface in detail and to compare its sex estimation accuracy using the geometric morphometric (GM) and traditional methodological approach. Our sample consisted of 121 specimens from 3 European osteological collections. The GM part of the study was based on 2D sliding semilandmarks that covered the outline of the auricular surface. Furthermore, several linear measurements and visual features (e.g. auricular surface elevation, postauricular sulcus) were chosen to test sex estimation accuracy using support vector machines. Concerning the GM analysis, the most notable sexual differences in the auricular surface outline relate to size. The best accuracy was achieved using form variables reaching 81.0%. Comparable accuracy (80.2%) was achieved using the metric approach, but combined with visual features the accuracy was increased to 93.4%. The GM approach was not very efficient in sexing the auricular surface outline, but the combination of visual features from the posterior ilium and metric variables of the auricular surface could be useful in sex estimation. Therefore, we provide a further testable linear discriminant equation based on this combination of variables.