Communication à un colloque
Résumé : In forensic contexts, highly fragmented human skeletons are frequently recovered either on domestic and terrorism cases, mass disasters or humanitarian international investigations. In front of such fragmented remains, forensic anthropologists cannot rely on standard morphoscopic or metric methods to give a reliable sex identification. Sexual diagnosis is indeed hampered by the non-optimal preservation of crucial dimorphic skeletal elements like the pelvis or the skull. To handle these frequent scenarii by providing an alternative sexing tool, this on-going doctoral project proposes to investigate the sexual dimorphism of two of the sturdiest regions of the human skeleton: the cranial base and the inner ear. Through a metric approach developped on a large European sample of 611 documented skeletons aged from 6 to 101 years, a set of measurements on the occipital and temporal bones were examined. Additionally, 93 CT-scans images were acquired to explore sexual variations in the inner ear of adults and subadults. Respective predictive models for sex were then established through linear discriminant analysis and logistic regression, with a posterior probability threshold of 0,7. Cranial base measurements yielded up to 86.8% of cross-validated accuracy in adults and 73.5% in subadults whilst bony labyrinth models reached up to 79.1% in adults and 85% in subadults. Results of this project have strong potentialities in forensic anthropology by allowing sexual identification of fragmented subadult and adult remains, which constitutes an essential step in establishing the biological profile of the deceased, in narrowing down the possible identities and in subsequently facilitating the investigation resolution.