Bayesian Concepts of Age Estimation Using Chronological Changes of Medial Clavicular Epiphysis

Introduction: Age estimation of undocumented persons has become a concern, especially with the current global migration crisis. Many states are still using old techniques, such as clinical examination, which have been associated with a wide margin of error. In addition, the analytical techniques used for age estimation need to be updated in tandem with the new approaches. The goal of the present study was to assess the precision of Schmeling’s age estimation method using Bayesian concepts of analysis. Methodology: In the current study, the use of Bayesian analysis for age estimation employing a chronological process of fusion of the medial epiphysis was assessed. A Bayesian analysis is a new statistical approach which incorporates unknown parameters of data, such as observed age (prior estimate), and how the date is likely to align with the actual age of an individual (posterior probability). Chest digital X-rays of 3000 patients aged between 15 and 30 years were collected retrospectively, examined based on the appearance of ossification centers and the extent of fusion of the epiphysis of the inner end of the clavicle. After thorough selection criteria, only 1605 chest radiographs qualified for the study. The phase of maturity and fusion of the epiphysis of the medial clavicle were subsequently documented in chronological order in accordance with the criteria applied by the standard classification stages, as documented by several authors. The gathered data were analyzed using probit regression and Bayesian transition analysis to determine the degree of accuracy in determining age, especially as one transitions from one stage to the next. The tests were derived from R-program. Results: Of More than half (53.5%) of the radiographs used met the selection criteria. Out of the targeted samples, 52.2% (805) were males while 49.8% (800) were females. The mean and median age of the total 1605 samples selected was 22.70 ± 0.12 years and 23 years respectively. The Bayesian concept analysis observed an error of ± 1.2 years in determining the age of an individual using this approach. Conclusion: The Bayesian analysis thus proved more accurate in determining age in the living, especially in the transition of stages.