Statistical methods for the comparison of measurements derived from orthodontic imaging

Abstract

Biometric comparison procedures for dental imaging methods with continuous outcome were reviewed, mainly concentrating on assessment and comparison of accuracy and precision according to the study design. Univariate graphical and numerical representations of corresponding deviations were summarized to derive a 'check list' of minimum information necessary to compare the measurement methods. The methods reviewed in this investigation are illustrated by the comparison of conventional (radiographic) cephalometry versus assessment using the DigiGraph® in 50 female children. A paired t-test and the corresponding confidence interval approach were used to assess deviations in location of two imaging methods; the test procedures of Maloney/Rastogi, Hahn/Nelson, and Grubbs were surveyed as proposals for the comparison of precisions in paired data. The Krippendorff coefficient was used as an aggregate measure for method concordance. Since these methods can be performed by simple modification of standard options available in most statistical software packages, this review intends to enable dental researchers to choose the correct methods, and perform adequate data analysis and representation.