Age Estimation and Sex Determination Using Raman Spectra of Human Dentine

Abstract

Objectives: This study aimed to determine the sex and estimate the age of individuals by analysing Raman spectra obtained from extracted teeth dentine.

Material and methods: A total of 25 male and 26 female extracted teeth were collected, disinfected, and fixed in auto acrylate before being cut near the cervical margin. The inclusion criteria were non endodontically treated teeth with the cervical third of the tooth crown preserved. The exposed dentine was recorded using a Raman spectrometer, with 100 scans per dentine site in the spectral range of 3500 to 200 cm^-1 and a resolution of 4 cm^-1. Background-corrected, normalized, and decomposed spectra were subjected to principal component analysis using MATLAB 2010 (The MathWorks, Natick, MA, USA). To assess age differences, prominent vibrational bands were observed and statistically analyzed, while sex differentiation employed the T-test on acquired data, exploring vibrational bands with significant intensity differences.

Results: The obtained results revealed a segmentation of spectra in the male sample group based on the age of teeth at extraction, with an accuracy of age estimation at 7.0048 years. A similar segmentation was observed in the female sample group, with an accuracy of 9.3863 years.

Conclusion: The recorded spectra were organized into datasets and analyzed with principal component analysis, demonstrating the feasibility of sex determination and age estimation using Raman spectra of teeth. Differences in classification accuracy between sexes may be attributed to hormonally-mediated differences in the biochemical composition of dentine between males and females. The findings of this study can provide valuable insights into methods for forensic and anthropological applications.

Keywords: Age Determination by Teeth; Age estimation; Forensic odontology; MeSH Terms: Dentin; Principal component analysis; Raman Spectrum Analysis; Raman spectroscopy; Sex Determination by Skeleton; Sex determination.

Figure 1

Segmentation of spectra based on the age, females (red: 10-19, green: 20-29, dark blue: 30-39, turquoise: 40-49, pink: 50-59, yellow: 60-69) RMSEC- root mean square error of calibration; RMSSECV- root mean square error of cross validation; CV- cross validation; R2- coefficient of determination; (Cal, CV)- (Calibration, Cross Validation); Red line: regression; Green line: ideal predicted values

Figure 2

Segmentation of spectra based on the age, males (red: 10-19, green: 30-39, dark blue: 40-49, turquoise: 50-59, pink: 60-69) RMSEC- root mean square error of calibration; RMSSECV- root mean square error of cross validation; CV- cross validation; R2- coefficient of determination; (Cal, CV)- (Calibration, Cross Validation); Red line: regression; Green line: ideal predicted values

Figure 3

The mean Raman spectra of male and female teeth, the loadings of the first principal component and the spectrum of the difference between the male and female teeth. A.u. -arbitrary units; Stt- Student t test

Figure 4

Schematic differentiation obtained by PCA/PCR processes (Male/Female). PC Principal component

Figure 5

Schematic differentiation obtained by PCA/PCR processes (Male/Female). PC Principal component