The dry human skulls included in this study came from the osteological collection of Department of Anatomy, Histology and Anthropology (Vilnius University) that consisted of skeletal burials from the Bokštro gatve 6 (13th-15th century) and Subačiaus gatve 7 (17th century) cemeteries in Vilnius, Lithuania.
The aforementioned skeletal series counted in total 470 and 129 individuals, respectively. From them 69 (Bokštro gatve 6, 42 males and 27 females) and 63 (Subačiaus gatve 7, 44 males and 19 females) were selected in accordance with following criteria:
- well stage of preservation (>75% skeleton, including intact or at least finely preserved skull with available cranial base and vault),
- age at death estimated as adultus or more,
- biological sex possible to estimate on the basis of skeletal traits,
- no cranial trauma, malformation or significant taphonomical fractures of skull,
- majority of cranial vault sutures visible.
We aim to develop a machine learning application to support future sex estimation based on the cranium. Low volume of the dataset and lack of the objective labels make it hard to evaluate with numeric quality-based metrics, what is the reason behind conducting following user studies.
Researchers:
- Dr Agata Bisiecka – anthropology specialist, data agregation, Department of Normal Anatomy, Pomeranian Medical University in Szczecin
- Mgr inż. Weronika Borek-Marciniec – machine learning specialist, co-author of the survey application, Department of Systems and Computer Networks, Faculty of Computer Science and Telecommunications, Wrocław University of Science and Technology
- Dr inż. Paweł Zyblewski – machine learning specialist, Department of Systems and Computer Networks, Faculty of Computer Science and Telecommunications, Wrocław University of Science and Technology
- Mgr inż. Szymon Wojciechowski – author of the survey application, Department of Systems and Computer Networks, Faculty of Computer Science and Telecommunications, Wrocław University of Science and Technology