Morphometrics for cephalometric diagnosis

This article demonstrates morphometric methods by applying them to an orthodontic sample. A total of 150 pretreatment cephalograms of consecutive patients (84 female, 66 male) were traced and digitized. Fifteen points were used for the analysis. The tracings were superimposed by the Procrustes method, and shape variability was assessed by principal component analysis. Approximately 70% of the total sample variability was incorporated in the first 5 principal components. The most significant principal component, accounting for 29% of shape variability, was the divergence of skeletal pattern; the second principal component, accounting for 20% of shape variability, was the anteroposterior maxillary relationship. It is recommended that Procrustes superimposition and principal component analysis be incorporated into routine cephalometric analysis for more valid and comprehensive shape assessment.