Musilová Barbora, Dupej Ján, Brůžek Jaroslav, Bejdová Šárka, Velemínská Jana
Department of Anthropology and Human Genetics, Faculty of Science, Charles University, Czech Republic.
Department of Anthropology and Human Genetics, Faculty of Science, Charles University, Czech Republic; Department of Software and Computer Science, Faculty of Mathematics and Physics, Charles University, Czech Republic.
Forensic Sci Int. 2019 Apr;297:364-369. doi: 10.1016/j.forsciint.2019.02.034. Epub 2019 Feb 26.
Assessing sex and population affinity is an important part of the process of biologically identifying unknown human remains, and the skull is usually one of the best structures for assessing both these components of the biological profile. Population affinity is known to be a hugely important variable when estimating sex because the manifestation of sexually dimorphic traits, body size or social and behavioural habits differs across populations. Therefore, for forensic purposes, the estimation of ancestry is a necessary step in the identification of bone remains. The present study improves on the results of a previously developed virtual method using the exocranial surface for sex estimation and assessing population affinity. The ability to assess these components of the biological profile was successfully tested on 208 individuals from two recent European populations. The original classifier was based on geometric morphometric analyses (CPD-DCA, PCA, SVM) and was able to assess the sex of individuals belonging to one French population with an accuracy exceeding 90 % Musilová et al. [1]. To improve the reliability of the method, the Czech population sample was added to the dataset, yielding the highest accuracy of 96.2 %; using the combined dataset, the reliability of the method was 91.8 %. Secondly, we used the same method utilizing inter-population differences to classify individuals based on the shape of the skull. The greatest accuracy rate was 92.8 %, which makes our method a promising tool for sex estimation and assessing population affinity.
评估性别和人群亲缘关系是生物学鉴定未知人类遗骸过程中的重要环节,而头骨通常是评估生物特征这两个组成部分的最佳结构之一。众所周知,在估计性别时,人群亲缘关系是一个极其重要的变量,因为两性异形特征、体型或社会及行为习惯在不同人群中的表现有所不同。因此,出于法医鉴定目的,估计祖先来源是骨骼遗骸鉴定中的必要步骤。本研究改进了先前开发的一种虚拟方法的结果,该方法利用颅外表面进行性别估计和评估人群亲缘关系。在来自两个近代欧洲人群的208名个体上成功测试了评估生物特征这些组成部分的能力。原始分类器基于几何形态计量分析(CPD-DCA、PCA、SVM),能够以超过90%的准确率评估属于一个法国人群的个体的性别(穆西洛娃等人[1])。为提高该方法的可靠性,将捷克人群样本添加到数据集中,准确率最高达到96.2%;使用合并后的数据集,该方法的可靠性为91.8%。其次,我们使用相同方法利用人群间差异根据头骨形状对个体进行分类。最高准确率为92.8%,这使我们的方法成为性别估计和评估人群亲缘关系的一种有前景的工具。