VIRTUAL ANTHROPOLOGY AND COMMINGLED SKELETAL REMAINS. RESOURCES AND POTENTIAL FOR PERSONAL IDENTIFICATION IN HUMANITARIAN FORENSIC CONTEXTS

Commingled skeletal remains represent a compelling topic in forensic anthropology, especially in view of the recent mass disasters in the Mediterranean Sea related to the migration crisis. With over 1,000 alleged victims, and thousands of commingled skeletal remains, the shipwreck of April, 8th, 2015 stands among these recurrent tragedies. The advancements of imaging techniques and 3D modelling have brought about innovative tools to forensic anthropology and paved the way to novel research avenues for the analysis of skeletal remains and for personal identification both of the living and the dead, as well. Foremost, research has extensively explored the quantification of similarities and differences between 3D models of anatomical structures as a proxy to infer relationships between AM e PM records and between bones of an individual. The general assumption is that the lowest is the distance between two superimposed models, the most likely is the possibility that the models share a common origin, hence they belong to the same individual. This dissertation delves into these innovative approaches utilizing the 3D-3D superimposition of virtual bone models to facilitate the sorting of commingled remains, ultimately seeking to comprehend the potential practical application in real case scenarios. The structure of this work follows the two main steps for the re- association of commingled bones, i.e., the joint-matching and the pair-matching, resulting in four research lines. The 3D models were produced with several acquisition techniques (laser scanning, sterephotogrammetry, segmentation and surface scanning) from bones of the CAL (Collezione Antropologica LABANOF) skeletal collection, from CT images of a hospital database and from the skeletal remains of the April, 18th, 2015 shipwreck. The 3D-3D joint-matching test on atlases and axes provided an exclusionary method and brought into question the validity of this method to re-associate planar joints. The 3D-3D pair-matching of iliac and of pubic bones further demonstrated that the method can be applied to portions of bones, as long as the extent of fragmentation preserves comparable structures. Indeed, the systematic test on the effect of postmortem damage on the pair-matching verified that 3D-3D superimposition is heavily impaired by the incompleteness of the remains. This study therefore confirmed with novel evidence the major setback that postmortem damage creates for any anthropological analysis. The application of the technique to a set of scapulae from the commingled remains of the shipwreck of April 18th, 2015, showed that 3D- 3D superimposition may represent a beneficial tool for the resolution of the commingling. This PhD project investigated unaddressed questions about 3D-3D superimposition applied to commingled remains. Indeed, the aim was not to replace visual and osteometric comparisons but to explore the possible application of an additional tool to be used together with traditional methods to strengthen the results within a multidisciplinary approach. Future research would better enlighten the role of up-to- date virtual techniques in the resolution of commingled assemblages from humanitarian contexts.