背景:快速的将心脏按其特征进行聚类可为后续统计分析和研究带来很大的便利.系统聚类法是将样品或变量按照其性质上的亲疏相似程度进行分类的一种多元统计方法.目的:提出用主成分一聚类分析的方法来描述心脏形态学形状并进行分类,对中国健康成年人的心脏X射线测量的各项指标进行综合评价.方法:搜集了36例健康成年人的胸片,并用MxLiteView软件手动测量了每幅胸片中代表心脏形态学形状常用的10个指标,用Matlab软件对测量指标进行主成分分析,然后对提取出的主成分进行聚类.结果与结论:主成分分析后提取出3个主成分变量,将36例样本用提取的主成分进行聚类,可将样本分为3类,分别代表了心脏的3类不同的心型.用该方法对心脏形态学形状进行快速分类,对心脏的统计和分类研究提供了一定的参考价值.-Background: The heart of their fast clustering features can bring great convenience for subsequent statistical analysis and research system clustering method is variable according to the sample or the nature of the closeness of the similarity of a multivariate statistical classification. methods Objective: proposed using principal component analysis of a clustering method to describe the shape of cardiac morphology and classification of the indicators Chinese healthy adult heart X-ray measurement of comprehensive evaluation methods: collected 36 cases of healthy adult person s chest and measuring MxLiteView software manual chest represents the heart of each piece of morphological shape commonly used 10 indicators to measure using Matlab software principal component analysis and principal components of the extracted clustering results aND CONCLUSION: after principal component analysis to extract the three main components of the variables, the 36 cases with samples extracted principal co