Subadult sex estimation

Forensic anthropologists are routinely asked to estimate a biological profile (i.e., age, sex, ancestry and stature) from a set of unidentified remains. In contrast to the abundance of collections and techniques associated with adult skeletons, there is a paucity of modern, documented subadult skeletal material, which limits the creation and validation of appropriate forensic standards. Many are forced to use antiquated methods derived from small sample sizes, which given documented secular changes in the growth and development of children, are not appropriate for application in the medico-legal setting. Therefore, the aim of this project is to use multi-slice computed tomography (MSCT) data from a large, diverse sample of modern subadults to develop new methods to estimate subadult age and sex for practical forensic applications. The research sample will consist of over 1,500 full-body MSCT scans of modern subadult individuals (aged birth to 20 years) obtained from two U.S. medical examiner's offices. Statistical analysis of epiphyseal union scores, long bone osteometrics, and os coxae landmark data will be used to develop modern subadult age and sex estimation standards. This project will result in a database of information gathered from the MSCT scans, as well as the creation of modern, statistically rigorous standards for skeletal age and sex estimation in subadults. Furthermore, the research and methods developed in this project will be applicable to dry bone specimens, MSCT scans, and radiographic images, thus providing both tools and continued access to data for forensic practitioners in a variety of settings.

After attending this presentation, attendees will appreciate: (1) the benefits of a semi-automated protocol to conduct volumetric measurements of complex skeletal elements using tissue density-based voxel-processing of computed tomography scans; and, (2) the presence of sexual dimorphism in frontal sinus volume in subadults. This presentation will impact the forensic science community by advancing understanding of subadult frontal sinus development through the use of 3D reconstructed Multi-Slice Computed Tomography (MSCT) scans. This research will also reinforce the invaluable nature and varied applicability a contemporary virtual skeletal repository offers to both forensic anthropology and osteoarchaeological fields of research. Due to high probability in the recovery of the frontal bones within fragmented remains, comparison of frontal sinus morphology in antemortem versus postmortem radiographs are routinely conducted for personal identification in forensic casework. While sexual dimorphism in frontal sinus size has been widely reported in adults, there is currently no research signifying the age of onset of sexual dimorphism in morphometric analyses of the frontal sinuses in subadults. Accordingly, this study seeks to establish a temporal profile of frontal sinus development and determine the timing of sexual dimorphism of the frontal sinus volume in a contemporary population of Australian subadults, using a standardized 3D modeling protocol. A total of 89 cranial Digital Imaging and Communications in Medicine (DICOM) datasets (males: n=45, females: n=44) aged 6 years to 19 years were accessed from the Skeletal Biology and Forensic Anthropology Virtual Osteological Database and imported into an advanced visualization software to generate an isosurface model and segmented masks of the right and left frontal sinuses. 1 A semi-automated method for volumetric calculation was developed to produce a standardized protocol for frontal sinus segmentation based on material discriminatory voxel intensity values. Voxels with intensity values ranging from-1,000 to 150 Hounsfield Units (HU) were selected, corresponding to the tissue density of air, fibrous connective tissue, cellular debris, and mucous that may be contained within the sinus cavity. Frontal chord (Bregma-Nasion) measurements were also conducted to standardize the crania, accounting for individual variability in cephalic size. Intra-and inter-observer testing was conducted on three sample datasets obtained from randomly selected individuals of varying ages. Intra-observer error demonstrated high precision and consistency between repeated measurements with a mean percent Technical Error of Measurement (%TEM) of 3.15% and Intra-Class Correlation Coefficient (ICC) of 1.00 (Confidence Interval (CI)=0.992-1.00). Similarly, inter-observer reliability demonstrated a high degree of observer agreement despite varying anatomical and radiographic experience, with a mean %TEM of 1.26% and ICC of 0.99 (CI=0.991-1.00). Age and sex effects were analyzed using independent Student t-tests with relevant post-hoc tests.