摘要
目的：使用Procrusts标准化以及聚类分析、判别分析方法对安氏Ⅱ类错[牙合]颅面形态进行分析，初步探讨使用形态测量以及多元统计对错[牙合]进行分类的方法。方法：收集北京大学13腔医学院·口腔医院正畸科1997～2000年就诊的安氏Ⅱ类错[牙合]患者894名，通过影像测量及最小平方和重叠法（PS）进行聚类分析以及亚型划分，并利用判别分析判别聚类分析准确性，建立判别方程，为今后新样本的分类和治疗后样本颅面形态的变化研究打下基础。结果：将样本头颅侧位片进行Procrusts标准化后，使用聚类分析将其分出11个具有各自形态特征的类型。建立3个判别方程，交互验证准确率为80．17％。结论：Ps重叠法用于形态学分类具有一定的优势；利用聚类分析，可以对安氏Ⅱ类错[牙合]进行进一步的划分；对于不同类别，判别的准确率有所不同；判别方程的建立为今后新样本以及治疗后样本分型的研究打下基础。
Objective：Cluster and discriminant analysis of the morphological characters of Angle＇s Class Ⅱ malocclusion was performed using Procrusts standardization. The classification of craniofacial morphology with the help of morphometry and multivariate statistics was also discussed. Methods： A total of 894 class Ⅱ patients were collected from the orthodontic department, Peking University School and Hospital of Stomatology during 1997 -2000. Using Procrusts standardization and cluster analysis, the samples were divided into different groups, discrimination equations were then established. Results： The samples were divided into Ⅱ subdivisions by Procrusts superimposition and cluster analysis. Three discrimination equations were established. The accuracy rate of cross-validated grouped cases was 80. 17%. Conclusion： Procrusts standardization had certain advantages in morphological classification; cluster analysis could be used in classification of Angle class Ⅱ malocclusion; For different types, the differentiate rate was not the same; the discrimination equations was the foundation for future research.
出处
《北京大学学报（医学版）》
CAS
CSCD
北大核心
2009年第1期71-75,共5页
Journal of Peking University:Health Sciences
基金
国家高技术研究发展计划专项经费资助(2001AA111392-3)~~
关键词
错[牙合]
安氏Ⅱ类
聚类分析
判别分析
Malocclusion, Angle class Ⅱ
Cluster analysis
Discriminant analysis
