A method of age estimation using Raman microspectrometry imaging of the human dentin.

The aim of this work is to establish a correlation between ageing and Raman spectra imaging of human teeth. Raman spectrum of an analysed sample is characteristic of its chemical components. By exploring the different dentinal areas on each tooth, we sought to find a correlation between noticeable variability on Raman spectra and the stages of dentinal evolution with advancing age. A study on 30 teeth analysed by Raman microspectrometry was used to select predictors of age, and it allowed us to determine a regression formula, with age as a dependent variable. Our data were computed by means of a statistical method called partial least squares (PLS) regression. This method was specially adapted for this kind of situation, where we have numerous predictors and relatively few observations. The regression formula was performed on a test sample of four more teeth, and we obtained a correct age estimation, with a mean error of more or less than 5 years. The main benefits of this method was a minimal and non-destructive tooth preparation, which led to an efficient age prediction, for any age group.