目的:估计火化遗骸的性别是困难的。由于其密度,岩骨经常在火葬中幸存下来。Wahl在1980年代观察到侧角是性别二态的。先前的研究表明,将女性与男性分开的各种分界点,很难复制,也很难应用。我们要测试Wahl方法,并将其与新的基于地标的版本进行比较。
方法:在本研究中,我们使用Micro-CT扫描测量了奥地利青铜时代晚期35个火化岩骨的侧向角。计算了两种不同方法的技术测量误差,以使内部耳道几乎在中线(手动或基于地标的交点)相交。此外,性别是根据形态特征和度量测量进行估计的。该信息用于逻辑回归建模以定义我们样本中的截止点。
结果:测量的技术错误表明,与更直观的手动交叉相比,基于地标的方法更精确。观察者间和观察者内的误差较低,从而提高了可靠性。逻辑回归模型在我们的样本中产生了良好的结果(p=0.02,R2=0.38,准确度=0.8)。平均侧角与针对史前火化岩骨的研究相似。
结论:提出的基于地标的方法是精确的,快,并且可以很容易地应用,即使是没有经验的研究人员。横向角的大小似乎是特定于人群的,但也取决于所应用的方法。我们建议使用更精确的基于地标的方法。
Estimating the sex of cremated human remains is difficult. The petrous bone frequently survives the cremation due to its density. Wahl observed the lateral angle to be sexually dimorphic in the 1980s. Previous studies showed various cut-off points to separate females from males, which are hardly replicable and difficult to apply. We want to test the Wahl method and compare it to a new landmark-based version.
In this study, we measured the lateral angle of 35 cremated petrous bones from late bronze age Austria using micro-CT scans. Technical errors of measurement were calculated for two different methods to intersect the internal acoustic meatus virtually in the midline (manual or landmark-based intersection). Furthermore, sex was estimated based on morphological features and metric measurements. This information was used in logistic regression modeling to define a cut-off point in our sample.
The technical errors of measurement suggested that a landmark-based method was more precise in comparison to a manual intersection which was much more intuitive. Inter- and intra-observer errors were low which improved reliability. The logistic regression model produced good results in our sample (p = 0.02, R2 = 0.38, accuracy = 0.8). The mean lateral angle was similar to studies which focused on prehistoric cremated petrous bones.
The proposed landmark-based method was precise, quick, and could be easily applied, even by unexperienced researchers. The size of the lateral angle seemed to be population-specific but also dependent on the method applied. We recommend to use the proposed landmark-based method which is more precise.