Follow us
P240 – Single-Tooth Age Estimation by Integrating DNA Methylation with Hyperspectral Imaging Analysis
Teeth represent a valuable resource in forensic identification; however, conventional morphological methods for age estimation often lack precision, particularly when dealing with single tooth. To address this, we developed a multimodal approach integrating DNA methylation analysis with hyperspectral imaging (HSI), leveraging the complementary strengths of molecular precision and non‑destructive spectral characterization. A DNA methylation age model was first constructed using 80 dental specimens (aged 10–65 years). A multiplex SNaPshot assay targeting five age‑associated CpG sites achieved high accuracy (testing MAE = 4.56 years, R² = 0.82). To enhance practical applicability, hyperspectral imaging (HSI) was employed as a non‑destructive, reagent‑free auxiliary tool. Spectral data were acquired from four defined regions of interest (coronal enamel, cervical region, root cementum, and apical cementum) on a separate set of intact teeth (n = 80, aged 10–65 years). Among these regions, the apical cementum yielded the best performance (testing MAE = 6.01 years, R² = 0.79), demonstrating that HSI alone provides a universally applicable and non‑invasive approach for age estimation. Finally, to combine the precision of molecular markers with the robustness of spectral analysis, we integrated the DNA methylation model with HSI data from the optimal region (apical cementum) using the same dental specimens. The resulting fused model (Methy + Apical Cementum) achieved improved accuracy (training MAE = 3.25 years), indicating that the combination of methylation and hyperspectral data offers a more precise solution for age estimation from single tooth specimens.
Authors
- Lagabaiyila Zha (Central South University, China)
- Siqi Chen (Central South University, China)
- Lihua Hou (Central South University, China)
- Ying Liu (Central South University, China)
On the same topic
