Age estimation from the biometric information of hand bones: Development of new formulas

Abstract

IntroductionIn the judicial context of the age estimation of living individuals, a new method was recently proposed, based on the collection of biometric information on hand bones radiographs. The aim of this study was to apply this method to a large French sample to provide new tools for age estimation Materials and methodsThe study sample consisted of metacarpals and proximal phalanges measurements of 1003 individuals aged less than 21 years. This sample was divided into two subgroups 1–12 and 13–21 years as the age of 13 is a relevant legal threshold for most European countries. A quadratic discriminant analysis was performed to identify the group to which an individual was most likely to belong. Age estimation formulas were also constructed from linear models: for each subgroup and the total sample. ResultsThe belonging of an individual to the 1–12 or 13–21 subgroup was determined with a correct classification rate of 89.8%. Age estimation formulas became less precise with age, with a mean absolute error ranging between 11 and 21 months. ConclusionWe proposed a two-step procedure for age estimation: firstly, the identification of the age group to which the individual is most likely to belong, and secondly, the age estimation of this individual by applying the appropriate formula.