Periodic Reporting for period 1 - Human Jigsaw (The Human Jigsaw: Matching articulating skeletal elements from mass burials)
Período documentado: 2019-05-13 hasta 2021-05-12
The objectives of the Human Jigsaw can be summarized as follows:
RO1: Development of techniques for the accurate matching of the elements of the lower limbs (femur, tibia) and the pelvic bones (os coxae). This will be accomplished by pair-matching bilateral elements but also by matching adjoining bones by examining the congruence of the hip and knee joints.
RO2: Application of the techniques developed in RO1 to prehistoric assemblages from Greece in order to contribute to discussions regarding funerary practices.
An open access software, “The CSG Toolkit”, which had been previously developed for extracting the cross-sectional geometric properties from long bones (Bertsatos and Chovalopoulou 2019) and has been uploaded to GitHub (https://github.com/pr0m1th3as/long-bone-diaphyseal-CSG-Toolkit) has been used in the context of the Human Jigsaw for the development of a novel method for pairing bilateral elements (Figures 1-2. The results were published in a paper titled “Advances in Osteometric Sorting: Utilizing Diaphyseal CSG Properties for Lower Limb Skeletal Pair-Matching” in the Journal of Forensic Sciences (https://zenodo.org/record/4743201). In addition, the open access software used for the development of this novel method was presented at the 2nd International Caparica Conference in Translational Forensics 2019.
The ongoing development of a novel sophisticated method for the accurate matching of the elements of the knee joint is anticipated to be completed by the end of 2021 and the software derivatives of this work will be published as open-source under the GNU General Public License v3.0 and disseminated through the project’s website. Furthermore, 59 skeletal elements from the late Roman necropolis of Mavropigi (northern Greece) housed in the Laboratory of Physical Anthropology, DUTH, were digitized and will be analysed using the new techniques in early 2022.
The Human Jigsaw as well as broader bioarchaeological and forensic anthropological topics were presented at the annual Researcher's Night (2019, 2020). Furthermore, several topics on physical anthropology were prepared as educational materials and presented to undergraduate students of the Democritus University of Thrace, Greece. This material is freely available to the public at the website: https://www.physicalanthropology.gr/. This website also contains a section dedicated to disseminating the methods developed in the context of the Human Jigsaw.
The Human Jigsaw involved prolonged data collection across major European human skeletal collections, during which the fellow had the opportunity to expand her network and familiarize with international procedures in the curation of skeletal remains. In addition, this project greatly expanded the fellow’s knowledge in machine learning and computer vision, and strengthened her skills in 3D documentation and geometric morphometric analysis. This skill set played an important role in the fellow securing a follow-up 3-year grant by the Hellenic Foundation for Research and Innovation (H.F.R.I) for her project RecHumS.
The Human Jigsaw scientific results are expected to have important implications in bioarchaeology and forensic anthropology since multiple/mass disturbed burials are very common in both disciplines and pose distinct interpretational challenges. In bioarchaeological contexts, the Human Jigsaw methodology will allow the identification of the number of individuals, which often employs robust elements such as parts of the femur, but also estimation of the age and sex of these individuals, which is principally based on the os coxae. Hence, this novel methodology will allow much more accurate palaeodemographic reconstructions and a better understanding of mortuary rituals and how they may have been differentiated across different age and sex groups. In forensic anthropological contexts, the Human Jigsaw methodology will facilitate the identification of the unknown subjects by allowing a very important initial sorting of the remains before destructive analyses, such as ancient DNA, are applied. In addition, the methodology developed by the Human Jigsaw is anticipated to pave the way for similar studies in zooarchaeology, where almost all assemblages are commingled and fragmentary, and the epiphyses are often the only well-preserved part of the bones. By adapting the machine learning algorithms to be used in ovicaprids, swine and cows, key archaeological information pertaining to animal management and consumption practices may be obtained. Finally, a highly accurate surface analysis method for matching adjoining bone surfaces can be of great value to prosthetic implants manufacturers. In order to facilitate further research applications, all algorithms produced by this project will be made openly available at GitHub, as already done with the CSG Toolkit and the pairing algorithms.