Estimation of facial morphology for forensic facial reconstruction using hand-derived morphometric measurements

Forensic identification is a fundamental process that plays a vital role in the identification of unidentified individuals as a result of natural disasters, mass deaths, accidents, homicides and man-made disasters. Especially in cases where the integrity of the body is compromised and only some body parts can be recovered, the applicability of classical methods decreases and alternative scientific approaches are needed. Although primary identification methods such as fingerprints, DNA analysis and dental records are generally considered the gold standard, these methods are not applicable in all cases. Therefore, the development of complementary techniques based on secondary methods is important for the fields of forensic medicine and anthropology. This thesis aims to make dimensional estimates of the individual's face and the regional anatomical structures that make up the face based only on the available hand morphometric data. For this purpose, 100 volunteer individuals (50 males, 50 females) between the ages of 19-24, without any congenital anomaly or trauma history in the hand and facial structure were included in the study. Photographs of the participants' dominant hands and neutral facial expressions in accordance with biometric standards were taken, and a total of 83 anthropometric measurements were recorded on the images obtained. These measurements were performed digitally using open-source ImageJ software and all data were collected by the same researcher to ensure standardization in measurements. In the study, 41 measurement parameters from the hand region (hand length, finger lengths, joint spans, inter-finger angles, etc.) were evaluated as independent variables, and 42 measurements from the face, eye, nose and lip regions were evaluated as dependent variables. Multiple linear regression analyses were performed for each facial region; the validity of the models was tested with analysis of variance (ANOVA) and their explanatory power was demonstrated with adjusted R² values. In addition, discriminant analysis method was used to reveal morphological differences due to sex, xviii and statistically significant differences were observed between male and female individuals. As a result of the analyses, significant correlations were found between hand measurements and facial structure and it was determined that some facial criteria can be predicted with high accuracy based on hand morphometry. These findings suggest that structural predictions can be made about an individual's face based on hand morphometry alone and that this method can make significant contributions to the identification process. The study also provides a new set of parameters for forensic facial reconstruction applications and demonstrates that not only the skull but also other postcranial structures such as the hand can be used in face prediction. It is possible that the obtained models can be integrated into digital face reconstruction systems in the future. As a result, this thesis presents a scientifically based, low-cost, practical and feasible alternative approach to forensic identification processes and provides new insights in the fields of forensic anthropology, biometric analysis and forensic technology.