Bayesian calibration for forensic age estimation.

Area of Science:

• Forensic Medicine
• Biostatistics
• Radiology

Background:

• Forensic age estimation is crucial for immigration and identification purposes.
• Current methods often rely on dental samples and regression models.
• Existing regression models can introduce bias, overestimating age in the young and underestimating in the old.

Purpose of the Study:

• To introduce a novel full Bayesian calibration method for forensic age estimation.
• To address the inherent biases in traditional regression-based age estimation techniques.
• To enhance the reliability and robustness of age assessment in forensic contexts.

Main Methods:

• Development of an asymmetric Laplace Bayesian calibration method.
• Utilizing the asymmetric Laplace distribution as the probability model.
• Comparison with existing Bayesian and classical age estimation methods using simulated data.

Main Results:

• The proposed asymmetric Laplace Bayesian calibration method demonstrated comparable accuracy to existing approaches.
• The new method proved significantly more reliable and robust, especially when the probability model was misspecified.
• The method was successfully applied to a real dataset of dental measurements from individuals of known age.

Conclusions:

• The asymmetric Laplace Bayesian calibration offers a more dependable approach to forensic age estimation.
• This method mitigates biases associated with traditional regression models.
• The findings support the utility of this novel Bayesian technique in forensic practice.