A volumetric approach to age estimation informed by voxel selection: Application to the spheno-occipital synchondrosis

Disparity regarding the age of spheno-occipital synchondrosal fusion is evident in anthropological literature, attributable to inconsistencies in visualization medium and scoring protocols. This chapter investigates quantification of voxel distributions embedded in tomography of the growth plate, as an alternative to modal, phase-based age estimation techniques. A semiautomated method for volumetric measurement was developed based on voxel intensity values of the Hounsfield scale (HU) that discriminate material composition of the synchondrosis in multislice computed tomography orthoslices. Nonlinear mixed-effects regression was utilized to formulate prediction models for age based on the cumulative proportion of voxels with intensity values 0–500 HU. Mean logistic models for predefined age cohorts are provided and their potential applicability for age estimation assessed using residual testing of an independent, modern Australian subsample. Residuals demonstrate high classification rates, based on the fit to the predictive models, with more precise age ranges than phase-based morphological standards derived for Australian subadults.