中国汉族人颅骨数字模型的性别判别方法

目的利用步进Fisher法建立多元性别判别函数,对西安地区成年人颅骨进行性别判别。方法从颅面数据库中选择西安地区133个（73男,60女）三维颅骨模型,利用计算机软件测量并统计颅骨的12项几何测量指标和2项角度测量指标,利用步进Fisher方法建立多元性别判别函数,并对方程进行回代检验。结果男性的12项几何测量指标均值大于女性,女性的2项角度测量指标均值大于男性,除2项指标（X7和X14）外,其他指标均存在显著性差异（P〈0.01）。由4项指标建立的判别方程,判别率男性为87.5%,女性为86.67%。结论利用计算机软件可以实现完整颅骨和无下颌的颅骨的性别判别。

Sex determination from digital skull model for the Han people in China

Objective This paper proposed a step Fisher to build multivariate linear discriminant function to achieve sex determination for adult skull of Xi＇an.Methods Firstly,133 digital skull samples（ male 73 and female 60） with sex information were chosen from craniofacial model database.Secondly,twelve geometric measurement indicators and two angle measurement indicators were computed separately.Finally,step fisher method was performed to build the sex dicriminant function and test samples were used to evaluate accuracy.Results Average of each geometric measurement indicator of male was bigger than those of female; conversely average of each angle measurement indicator of male was smaller than female.Except two indicators（ X7 and X14）,the other indicators existed significant difference（ P 0.01）.According to sex dicriminant function consisting of four indicators,the accuracy of male was 87.5% and female was 86.67% separately for complete skull.Conclusion Sex determination for both complete skull and maxilla can be estimated by using developed computer software.