Accuracy of the London Atlas and RP Atlas in age estimation of Southern Brazilians

To assess the accuracy of the London Atlas and RP Atlas methods estimating age and diagnosing adulthood in a Southern Brazilian population.

A total of 1,099 panoramic radiographs from individuals aged 15.00-22.99 years were analyzed. The London Atlas and RP Atlas tools were used to estimate age based on the developmental stages of the upper and lower left third molars (teeth 28 and 38). Statistical analysis included Spearman's rank test, Bland-Altman plots, Lin's concordance coefficient, and ROC curves, with adulthood (>18 years) as the cutoff (p<0.05).

Both methods showed high accuracy, with the London Atlas exhibiting lower mean errors. Errors increased with age and were higher in females. The best adulthood cutoff for the London Atlas was 18.5 years for tooth 28 (78.5% accuracy) and 19.5 years for tooth 38 (80.0% accuracy). For the RP Atlas, the cutoff was 17 years for both teeth (80.8% and 80.0% accuracy, respectively). Both methods were suitable for age estimation and adulthood diagnosis, with the London Atlas tending to overestimate age - an advantage in legal contexts requiring caution.

Validating age estimation methods in specific populations enhance accuracy, preventing errors when applying methods developed for other populations and ensuring suitability for the local context. Research that applies these methods can provide a foundation for adjustments to the Atlases and their application in diverse subpopulations, and even serve as the basis for the establishment of an international data repository.