Assessing the Greater Sciatic Notch With 2-D Shape Analysis for Sex Estimation

Abstract

A correct biological profile leads to a better understanding of the past and assists in the identification of human remains within bioarchaeology and forensic casework. Sex estimation forms a critical component of a biological profile. With the advancement of technologies such as geometric morphometrics (GMM), new methods and a deeper understanding of morphological features can be investigated digitally. However, how well do these methods compare to standard visual methods and how easy are they to employ? This research investigates the use of 2-D shape analysis and visual morphological methods for sex estimation using the greater sciatic notch (GSN). A total of 202 adult os coxae were photographed and analyzed from the Spitalfields Coffin Plate Collection housed at the Natural History Museum, UK. Each os coxae was analyzed digitally to extract a “line” for elliptical fourier analysis (EFA) and subsequent discriminant function analysis (DFA). Os coxae were also scored using two well established morphological methods for the GSN. This study found an overall accuracy of 82.81% when using the computational method (EFA and DFA). Lower accuracies were found for the visual methods with the Bruzek method correctly classifying 82.17% and the Walker method resulting in a much lower accuracy at 72.77%. The finding of this study showcases the benefits of using more computational methods such as shape analysis/GMM. However, it has a nearly identical overall error rate to the Bruzek method and higher accuracy than the Walker method and therefore is a suitable and accurate method for sex estimation. As these practices are evolving, practitioners will have to balance the cost/benefit (e.g., time, training, and accuracy) of using the different techniques while continuing to refine and combine approaches for optimal results in biological profiling.