Estimation of stature and sex from pelvic measurements in a Chinese population

Abstract

ABSTRACT Individual identification is one of the most challenging aspects of forensic science. The aim of this study was to develop population-specific discriminant function equations and stature prediction equations for predicting sex and stature from measurements of the pelvis in a contemporary Chinese population by using 3D CT (three-dimensional computed tomography) pelvic images. The stepwise analysis of all 13 measurements yielded a sex classification accuracy rate of 99.5 % and a sex bias of 0.1 %. The sub-pubic angle is a reasonably accurate single parameter, with the accuracy of 97.4 % in sex determination. Stature was estimated using the equations that involved the dimensions of a single variable. The accuracy of stature prediction from univariate analysis ranged from 5.06 to 5.53 cm in males, and from 4.62 to 5.09 cm in females. Multiple regression equations were presented, with the accuracy of stature prediction being 4.51 cm for males and 4.22 cm for females. This article provides a reliable alternative for sex diagnosis from reconstructed 3D CT pelvic images and it could be effectively used in forensic cases for a Chinese population. Furthermore, the equations presented for stature estimation in this study should be used as alternatives in forensic cases when long bones were unavailable for stature estimation.