Abstract
Facial approximation is performed on unidentified skull remains when other methods of identification have failed. The generated facial image is presented to police departments and media to
promote public awareness and in turn facilitate recognition. The reliability of a facial approximation is dependent on the skull preservation. When the mandible is absent, practitioners might consider a reliable approximation to not be possible and often refrain from attempting a reconstruction.
This study aims to perform a validation study on a recently published 3D method for mandible estimation. Moreover, a new population-specific approach on the target sample is tested.
The sample comprised CT scans of 50 white North American individuals (25 males, 25 females) taken from The New Mexico Decedent Image Database. Skull scans demonstrating no signs of pathology or trauma were selected. 54 measurements were collected from each cranium, and 24 measurements from each mandible following Babacan et al.’s (2021) method. Error rates produced by
the existing method were calculated and a regression analysis was performed as a pilot study to test whether population-specific formulae provided higher accuracy rates.
The results obtained using Babacan et al. (2021) method on the sample under study demonstrated, on average, an inaccuracy of 8 millimetres and a bias of 4 millimetres, with similar values obtained for males and females, separately. Regression analysis produced a range of simple and multiple models with errors comparable or lower than Babacan et al. (2012) errors.
Our results demonstrated differences between the reference population and the target population that might be due to intrinsic and extrinsic factors such as genetics and environmental adaptations. The need of a population-specific method is further suggested. Future research should involve the validation of other existing methods on mandible estimation, as well as a larger and more
heterogeneous sample to confirm and validate our preliminary population-specific formulae.
promote public awareness and in turn facilitate recognition. The reliability of a facial approximation is dependent on the skull preservation. When the mandible is absent, practitioners might consider a reliable approximation to not be possible and often refrain from attempting a reconstruction.
This study aims to perform a validation study on a recently published 3D method for mandible estimation. Moreover, a new population-specific approach on the target sample is tested.
The sample comprised CT scans of 50 white North American individuals (25 males, 25 females) taken from The New Mexico Decedent Image Database. Skull scans demonstrating no signs of pathology or trauma were selected. 54 measurements were collected from each cranium, and 24 measurements from each mandible following Babacan et al.’s (2021) method. Error rates produced by
the existing method were calculated and a regression analysis was performed as a pilot study to test whether population-specific formulae provided higher accuracy rates.
The results obtained using Babacan et al. (2021) method on the sample under study demonstrated, on average, an inaccuracy of 8 millimetres and a bias of 4 millimetres, with similar values obtained for males and females, separately. Regression analysis produced a range of simple and multiple models with errors comparable or lower than Babacan et al. (2012) errors.
Our results demonstrated differences between the reference population and the target population that might be due to intrinsic and extrinsic factors such as genetics and environmental adaptations. The need of a population-specific method is further suggested. Future research should involve the validation of other existing methods on mandible estimation, as well as a larger and more
heterogeneous sample to confirm and validate our preliminary population-specific formulae.
| Original language | English |
|---|---|
| Publication status | Published - 1 Oct 2022 |
| Event | Annual Symposium 2022, Forensic Anthropology Society of Europe (FACE) - University of Crete, Heraklion, Greece Duration: 1 Oct 2022 → 1 Oct 2022 http://forensic.med.uoc.gr/fase2020/symposium.jsp |
Conference
| Conference | Annual Symposium 2022, Forensic Anthropology Society of Europe (FACE) |
|---|---|
| Country/Territory | Greece |
| City | Heraklion |
| Period | 1/10/22 → 1/10/22 |
| Internet address |