Improving the concordance between various anteroposterior cephalometric measurements using Procrustes analysis

The aim of this study was to investigate a method which minimizes the effects of geometric distortion on various cephalometric measurements used to determine sagittal discrepancy, such as ANB angle, Wits appraisal, AB plane angle, projections on the palatal plane, Frankfort horizontal (FH) plane, the mandibulomaxillary bisector, and the SN line, in an attempt to optimize the correlation between them. This was accomplished by superimposing the Bolton 12-year male-female averaged template on a patient's tracing using Procrustes analysis and performing measurements while exchanging the patient's reference landmarks/planes (point N, the mandibulomaxillary bisector, FH plane, occlusal plane, palatal plane, and SN line) with those of the template. The normalized measurements were then compared with their classic counterparts using correlation coefficients. The above cephalometric analyses, classic and normalized, were applied to 71 patients [26 males: mean age 13.1 years, standard deviation (SD) 1.1 years and 45 females: mean age of 14.6 years, SD 8.2 years]. Spearman's rank correlation coefficient was calculated between the classic measurements and their normalized counterparts, resulting in a consistent increase in the correlation between the normalized measurements in comparison with the classic ones. This increase varied in absolute value from 0.052 to 0.405. All normalized measurements were highly correlated (P > 0.742, absolute value). Although correlation calculations do not represent a true measure of diagnostic performance, it is hoped that improving their correspondence heightens the possibility of the different tests agreeing on the patient's sagittal discrepancy, decreasing the possibility of differing, or even totally opposing diagnostic outcomes resulting from their application to (clear-cut) Class I, II, and III patients.