Facial soft tissue thickness and facial recognition in black South African adults

Abstract

Forensic facial approximations are used when conventional methods of identification provide insufficient results. The success of these facial approximations are founded on access to the applicable facial soft tissue thicknesses (FSTT), the culmination of muscle and connective tissue extending the distance from the outer surface of the bone to the external extreme of the skin. The aims of this study were to supplement existing South African FSTT data and investigate their reliability by utilizing facial recognition tests. Sixty computed tomography (CT) scans were obtained retrospectively from the Wits Donald Gordon Medical Centre (Johannesburg) after obtaining ethical clearance. On each scan, FSTTs were measured at 10 midline and 10 bilateral landmarks. South African sex-specific & pooled data, as well as FSTT from the C-Table repository (available through http://www.CRANIOFACIALidentification.com) were utilized to perform several facial approximations using the Manchester method. Assessors were then asked to study the facial arrays and indicate which photographs of the 20 presented to them resembled the facial approximations. The FSTT measurements were collected for 47 females (78.3%) and 13 males (21.7%). The range of the midline and lateral means were 2.4 to 11.3 mm and 6.5 to 29.8 mm respectively. The range of the midline and lateral shorths were 1.8 to 10.4 mm and 5.6 to 27.5 mm respectively. Moreover, the 75-shormaxes of midline and lateral landmarks ranged from 3.3 to 13.2 mm and from 8.4 to 37.6 mm respectively. Statistical significance between sexes for FSTT, was found to only exist at the mid-zygomatic landmark (t-test; p<0.05). One significant difference was accounted for at the supra M2 pertaining to bilateral asymmetry. Age-related differences, mostly in the vicinity of the eyes and nose, were however, statistically significant at several midline and bilateral landmarks. The facial approximations produced in this study displayed subtle changes depending on the data used to achieve them (i.e.: sex-specific or pooled data). Variations in the appearance of the jaw line, parotid region and mouth were the most noticeable; however, these did not influence recognisability. All categories of assessors exhibited above-chance success during the facial array tests. The female approximations were paired with correct photograph 30% more frequently than the male approximations; however, the difference was not statistically significant. White females with experience of anatomy and/or art displayed the highest success during the facial arrays and indicated that the bridge of the nose and breadth of the nostrils were the most telling features. The results of this study have uncovered that the categorization of FSTTs does not significantly influence the recognition rates of facial approximation. Facial approximations produced using pooled data were identified more easily. Furthermore, recognition biases associated with sex and ancestry were not significant in the current South African study.