Jean-Luc GREMAUD
Forensic Science School, University of Lausanne, Lausanne-Dorigny, Switzerland
Forensic Science Unit, Police of Valais, Sion, Switzerland

The use of dental health data to identify deceased persons has been tried and tested for a number of years now. The first part of this article reviews the main research projects which have been conducted with radiographic and nonradiographic dental health data. Those studies have examined two specific factors: the intra- and intervariability of dental health data. These two factors are combined into a Bayesian framework. Two alternative hypotheses are compared, 1) the identity of a deceased person corresponds to a presumptive identity, and 2) the identity of the deceased person corresponds to someone else. According to this approach, the results of dental analyses comparing post-mortem (PM) data with ante-mortem (AM) data from a presumptive source are studied on the basis of a likelihood ratio (LR). The ratio allows for an assessment of the value of the compared dental data for the purpose of identifying the deceased person. In order to assess dental intravariability, a concept known as “validity” is defined as an indicator of conditions governing intravariable data comparison.

Forensic sciences; Odontology; Identification; Statistics; Bayes; Likelihood ratio.